The present invention concerns antibody-like binding protein specifically binding to CD3 and CD123. The invention also relates to pharmaceutical compositions comprising said antibody-like binding protein and the use of said pharmaceutical compositions and antibody-like binding protein to treat cancer. The invention further relates to isolated nucleic acids, vectors and host cells comprising a sequence encoding said antibody-like binding protein.
The first generation of bispecific antibodies was developed over 20 years ago. Since then a number of clinical studies have tested bispecific antibodies engineered to target cancer cell surface antigens. This group of anti-cancer fusion proteins contains two or more functional domains that localize immunological effector cells in the proximity of targeted cancer cells to achieve anti-cancer activity.

As bispecific antibody technology developed, a different group of fusion proteins named bispecific T-cell engagers (BiTE) were generated by connecting two antibody single chain variable regions (scFv) only (no Fc amino acid segments were included) with a flexible linker, one scFv binds targeted cells and the other binds CD3 on T cell surface. One BiTE, blinatumomab, with CD19xCD3 bi-specific binding activities showed promising results in Phase II clinical trials for patients with minimal residual disease in B-lineage acute lymphoblastic.

CD123 (the interleukin-3 receptor alpha chain IL-3Ra) is a tumor antigen over-expressed in a variety of hematological neoplasms. The majority of AML blasts express surface CD123 and this expression does not vary by subtype of AML. Higher expression of CD123 on AML at diagnosis has been reported to be associated with poorer prognosis. It has been reported that CD123 is expressed on leukemic stem cells (LSCs). There is growing evidence to suggest that AML arises from these leukemic stem cells (LSCs) which have been shown to be quiescent and relatively resistant to DNA damaging chemotherapy. The increased expression of CD123 on LSCs compared with hematopoietic stem cells (HSCs) presents thus an opportunity for therapeutic targeting of AML-LSCs.

The monoclonal antibody (MAb) 7G3, raised against CD123, has previously been shown to inhibit IL-3 mediated proliferation and activation of both leukemic cell lines and primary cells (US Patent No. 6,177,078). However, it has remained unclear whether targeting CD123 can functionally impair AML-LSCs.

The use of CD123xCD3 antibody-like binding protein leads to tumor cell killing, as herein shown by the inventors.

The idea of producing a bispecific antibody-like binding protein with CD123xCD3 bi-specific binding activities has already been proposed and described in the international patent application WO2013/173820.

Furthermore, a CD123 x CD3 Dual Affinity Re-Targeting (DART) Bi-Specific Antibody Based Molecule from MacroGenics entered phase I clinical trials in 2014.

However, as shown by the inventors, the CD123xCD3 Dual Affinity Re-Targeting (DART) Bi-Specific Antibody Based Molecule from MacroGenics, for example, has an activation of 82% of CD4+ expressing T-cells and 83% of CD8+ expressing T-cells in the absence of target cells. The inappropriate activation of T-cells may lead to severe side effects, such as the cytokine release syndrome. The cytokine release syndrome refers to the release of cytokines by the activated T cells producing a type of systemic inflammatory response similar to that found in severe infections and characterized by hypotension, pyrexia and rigors. Deaths due to cytokine release syndrome have been reported for example for the anti-CD3 antibody OKT3.

Anti-CD3/anti-CD123 antibody-like binding proteins are described in patent application n° PCT/EP2016/051386 which was not yet published at the priority filing date of the instant patent application (article 54(3) according to European Patent Convention). Therefore, in spite of these advancements in bispecific antibody technology, there remains a need for additional cancer therapeutics, particularly those that efficiently target and kill cancer cells, either directly or indirectly. Moreover, there is a need to develop new anti-CD3/anti-CD123 antibody-like binding proteins having the desired biological activity, good metabolic, pharmacokinetic and safety profile, and also, that can be manufactured in large scale compatible with industrial practice.

Accordingly, in context of the present invention, the inventors succeeded in developing several variants of anti-CD3/anti-CD123 antibody-like binding proteins containing mutations, such as a RF mutation and Knob-into-hole mutations, thereby reducing the aggregation of said anti-CD3/anti-CD123 antibody-like binding proteins during expression. By reducing the amount of aggregates, an increased amount of heterodimer of the antibody-like binding proteins during expression and purification can be achieved, thus increasing the yield of the purified anti-CD3/anti-CD123 antibody-like binding proteins.

The present invention thus refers to anti-CD3/anti-CD123 antibody-like binding proteins comprising mutations leading to reduced aggregation during expression and/or purification. Said anti-CD3/anti-CD123 antibody-like binding proteins have a low T-cell activation capacity in the absence of CD123 expressing target cells, such as THP-1 cells, but a high capacity of activation of T-cells in the presence of CD123 expressing target cells, such as THP-1 cells.

Anti-CD3 antibodies

"CD3" denotes an antigen that is expressed on T-cells as part of the multimolecular T-cell receptor complex and that consists of at least three different chains CD3e, CD35 and CD3y. CD35 and CD3y have a low sequence identity and/or similarity to human CD3e (similarity and identity is less than 20%). CD3e and CDR35 can form together a complex, so called "CD3e/5-complex". CD3e also forms a complex with CDR3y, the so-called "CD3e/v-complex" Clustering of CD3 on T-cells, e.g., by immobilized anti-CD3-antibodies, leads to T-cell activation similar to the engagement of the T-cell receptor but independent from its clone typical specificity. "CD3e" comprises three domains, an intracellular domain, a transmembrane domain and an extracellular domain.

Most prior art anti-CD3-antibodies recognize the CD3e-chain. One of such prior art anti-CD3-antibodies is OKT3. Prior art has exemplified T cell activation events employing antibody molecules for example by employing the antibody molecule OKT3. The anti-CD3 antibody and variant thereof have been described in the prior art (US 4,361 ,549; US 4,361 ,549; US 5,885,573; US 5,929,212; and WO 98/52975 or US 5,955,358). OKT3 has been further used as potent immunosuppressive agent in clinical transplantation to treat allograft rejection (Thistlethwaite 1984, Transplantation 38, 695-701 ; Woodle 1991 , Transplantation 51 , 1207-1212; Choi 2001 , Eur. J. Immunol. 31 (1 ), 94-106).

Major drawbacks of this therapy are T cell activation manifested in cytokine release due to cross-linking between T cells and FcyR-bearing cells and the human anti-mouse antibody (HAMA) response. Several publications have described alterations such as humanization of OKT3 to reduce these side effects: US 5,929,212; US 5,885,573 and others. On the other hand, OKT3 or other anti-CD3-antibodies can be used as immunopotentiating agents to stimulate T cell activation and proliferation (US 6,406,696 Bluestone; US 6,143,297 Bluestone; US 6,1 13,901 Bluestone; Yannelly 1990, J. Immunol. Meth. 1 , 91 -100). Anti-CD3-antibodies have also been described as agents used in combination with anti-CD28-antibodies to induce T cell proliferation (US 6,352,694). OKT3 has further been used by itself or as a component of a bispecific antibody to target cytotoxic T cells to tumor cells or virus infected cells (Nitta 1990, Lancet 335, 368-376; Sanna 1995, Bio/Technology 13, 1221 -1224; WO 99/54440).

Approaches up to now using antibodies as agents for recruiting T-cells have been hampered by several findings. First, natural or engineered antibodies having a high binding affinity to T-cells often do not activate the T-cells to which they are bound. Second, natural or engineered antibodies having a low binding affinity to T-cells are also often ineffective with respect to their ability to trigger T-cell mediated cell lysis.

A reference sequence of full-length human CD3e protein, including the signal peptide, is available from the Uniprot database under accession number P07766 (as available on December 12, 2014) and herein enclosed under SEQ ID NO: 1.

A reference sequence of full-length Macaca fascicularis CD3e protein, including the signal peptide, is available from the Uniprot database under accession number Q95LI5 (as available on December 12, 2014) and herein enclosed under SEQ ID NO: 2.

A sequence of mature human CD3e His-tagged Fc-fusion proteins, cloned by the inventors from genomic DNA, is disclosed under SEQ ID NO: 3. Said mature human CD3e His-tagged Fc-fusion protein comprises amino acids 23 to 126 of the full-length human CD3e protein and thus comprises the extracellular domain of human CD3e.

A sequence of mature Macaca fascicularis CD3e Fc-fusion protein, cloned by the inventors from genomic DNA, is disclosed under SEQ ID NO: 4. Said mature Macaca fascicularis CD3e Fc-fusion protein comprises amino acids 23 to 1 17 of the full-length Macaca fascicularis CD3e protein and thus comprises the extracellular domain of human or Macaca fascicularis CD3e, containing one alanine to valine exchange at the amino acid position 35 in comparison to amino acid position 57 of the wild-type sequence.

Domain organization of human and Macaca fascicularis CD3e is as it follows (based on Uniprot P07766 sequence (human) and Uniprot Q95LI5 sequence {Macaca fascicularis)):

CD3e domains Positions on SEQ ID NO :1 Positions on SEQ ID NO : 2

(human) (Macaca fascicularis)

Extracellular 23 - 126 22 - 1 17

Transmembrane domain 127 - 152 1 18 - 138

Cytoplasmic 153 - 207 139 - 198

Accordingly, the extracellular domain of human CD3e consists of amino acids at positions 23 - 126 of SEQ ID NO: 1 and the extracellular domain of Macaca fascicularis CD3e consists of amino acids at positions 22 - 1 17 of SEQ ID NO: 2.

The humanized anti-CD3 antibody "hz20G6" of which the sequences of the heavy and light chain variable domains are used in context of the "hz20G6Xhz7G3" antibody-like binding proteins comprises

- a heavy chain variable domain consisting of sequence

QVQLVESGGGVVQPGRSLRLSCAASGFTFTKAWMHWVRQAPGKQLEWVAQIKD KSNSYATYYADSVKGRFTISRDDSKNTLYLQMNSLRAEDTAVYYCRGVYYALSPFDYW

GQGTLVTVSS

(SEQ ID NO: 9, with CDRs shown in bold characters) comprising CDR1 -H of sequence SEQ ID NO: 5, a CDR2-H of sequence SEQ ID NO: 6, and a CDR3-H of sequence SEQ ID NO: 7, and

- a light chain variable domain consisting of sequence

DIVMTQTPLSLSVTPGQPASISCKSSQSLVHNNANTYLSWYLQKPGQSPQSLIYKV SNRFSGVPDRFSGSGSGTDFTLKISRVEAEDVGVYYCGQGTQYPFTFGSGTKVEIK

(SEQ ID NO: 10, with CDRs shown in bold characters) comprising CDR1 -H of sequence SEQ ID NO: 1 1 , a CDR2-H of sequence 'KVS', and a CDR3-H of sequence SEQ ID NO: 8.

The humanized anti-CD3 antibody "hz20G6" used in context of the present invention displays high affinity for both human and Macaca fascicularis CD3 protein, and has however a low T-cell activation in the absence of target cells.

The anti-CD3 antibody "hz20G6" binds in particular to the extracellular domain of human CD3, or of both human and Macaca fascicularis CD3. More specifically, the antibody binds to CD3e. More specifically, the anti-CD3 antibody binds to the human and Macaca fascicularis extracellular domain of CD3e. The anti-CD3 antibody binds to CD3e when present in the form of a complex, such as a CD3e/5 complex, or when present as single protein, indifferently whether expressed in isolated form, or present in a soluble extracellular domain or full-length membrane-anchored CD3e as present in for example in T-cells.

The anti-CD3 antibody "hz20G6" used in context of the present invention is specific for the surface human CD3 protein, or of both human and Macaca fascicularis CD3 proteins, in particular to CD3e. In particular, the antibody does not bind to, or does not significantly cross-react with the extracellular domain of the aforementioned human and Macaca fascicularis CD3y and/or CD35 protein(s).

The anti-CD3 antibody "hz20G6" used in context of the present invention is the humanized version of the anti-CD3 antibody "20G6". The anti-CD3 antibody "20G6" has a ka of 3,5*104 (1/Ms), a kd of 2,7 *10"4 (1/s) resulting in a KD of 7,7*10"9 (M) to human CD3e/5 complexes and a ka of 2,7*104 (1/Ms), a kd of 2,2 *10"4 (1/s) resulting in a KD of 8,2*10"9(M) to Macaca fascicularis CD3e/5 complexes, both as measured by Biacore (data not shown). The anti-CD3 antibody "20G6" thus has a ratio of affinity for Macaca fascicularis CD3 on affinity for human CD3 (KD(Macaca fascicularis)/ KD(human)) which is 1 . The anti-CD3 antibody "20G6" and antibody-like binding proteins derived therefrom may thus be used in toxicological studies performed in monkeys the toxicity profile observed in monkeys relevant to anticipate potential adverse effects in humans

Accordingly, anti-CD3 antibody "20G6" used in context of the present antibody-like binding proteins has an affinity (KD) for human CD3 or Macaca fascicularis CD3, or both, which is < 10nM.

Anti-CD123 antibodies

"CD123" (Cluster of Differentiation 123) is also known as "Interleukin 3 receptor, alpha (IL3RA)" or "IL3 ", "IL3RX", "IL3RY", "IL3RAY", "hlL-3Ra" and denotes an interleukin 3 specific subunit of a heterodimeric cytokine receptor. The functional interleukin 3 receptor is a heterodimer that comprises a specific alpha chain (IL-3A; CD123) and the IL-3 receptor beta chain (βθ; CD 131 ) that is shared with the receptors for granulocyte macrophage colony stimulating factor (GM-CSF) and interleukin 5 (IL-5). CD123 is a type I integral transmembrane protein with a deduced Molecular Weight of about 43kDa containing an extracellular domain involved in IL-3 binding, a transmembrane domain and a short cytoplasmic tail of about 50 amino acids. The extracellular domain is composed of two regions: an N-terminal region of about 100 amino acids, the sequence of which exhibits similarity to equivalent regions of the GM-CSF and IL-5 receptor alpha-chains; and a region proximal to the transmembrane domain that contains four conserved cysteine residues and a WSXWS motif, common to other members of this cytokine receptor family. The IL-3 binding domain comprises about 200 amino acid residue cytokine receptor motifs (CRMs) made up of two Ig-like folding domains. The extracellular domain of CD123 is highly glycosylated, with N-glycosylation necessary for both ligand binding and receptor signaling. The protein family gathers three members: IL3RA (CD123A), CSF2RA and IL5RA. The overall structure is well conserved between the three members but sequence homologies are very low. One 300 amino-acid long isoform of CD123 has been discovered so far, but only on the RNA level which is accessible on the Getentry database under the accession number ACM241 16.1 .

A reference sequence of full-length human CD123 protein, including signal peptide, is available from the NCBI database under the accession number NP_002174.1 and under the Uniprot acession number P26951 and is herein disclosed under SEQ ID NO: 12 (as available on December 14, 2014).

A reference sequence of full-length Macaca fascicularis CD123 protein, including signal peptide, is available from GenBank database under the accession number EHH61867.1 and under the Uniprot acession number G8F3K3 and is herein disclosed under SEQ ID NO: 13 (as available on December 14, 2014).

A sequence of a mature human CD123 His-ll tagged Fc-fusion protein, cloned by the inventors from genomic DNA, is disclosed under SEQ ID NO: 14. Said mature human CD123 Fc-fusion protein comprises amino acids 19 to 305 of the full-length human CD123 protein and thus comprises the extracellular domain of human CD123.

A sequence of a mature Macaca fascicularis CD123 His-ll tagged Fc-fusion protein, cloned by the inventors from cDNA, is disclosed under SEQ ID NO: 15. Said mature Macaca fascicularis CD123 Fc-fusion protein comprises amino acids 19 to 305 of the full-length Macaca fascicularis CD123 protein and thus comprises the extracellular domain of Macaca fascicularis CD123.

Domain organization of human and Macaca fascicularis CD123 is as follows (based on the human CD123 sequence accessible in the NCBI database under accession NP_002174.1 (SEQ ID NO: 12) and based on the Macaca fascicularis CD123 sequence accessible in the Uniprot database under acession number G8F3K3, SEQ ID NO: 13):

Human CD123 Positions on SEQ Positions on SEQ ID NO:

domains ID NO: 12 13

(human) (Macaca fascicularis)

Extracellular 19 - 305 19 - 305

Transmembran 306 - 325 306 - 325

e domain

Cytoplasmic 326 - 378 326 - 378

Accordingly, the extracellular domain of human CD123 consists of amino acids at positions 19 - 305 of SEQ ID NO: 12.

CD123 (the interleukin-3 receptor alpha chain IL-3Ra) is a tumor antigen over-expressed in a variety of hematological neoplasms. The majority of AML blasts express surface CD123 and this expression does not vary by subtype of AML. Higher expression of CD123 on AML at diagnosis has been reported to be associated with poorer prognosis. CD123 expression has been reported in other hematological malignancies including myelodysplasia, systemic mastocytosis, blastic plasmacytoid dendritic cell neoplasm (BPDCN), ALL and hairy cell leukemia.

CD123 is expressed on AML leukemic stem cells and growing evidences suggest that AML arises from these LSCs, which have been shown to be quiescent and relatively resistant to DNA damaging chemotherapy. It is hypothesized that the persistence of LSCs underpins relapse after initial remission and thus the eradication of LSCs can be considered a requirement for cure, and an important therapeutic goal.

The monoclonal antibody (MAb) 7G3, raised against CD123, has previously been shown to inhibit IL-3 mediated proliferation and activation of both leukemic cell lines and primary cells (US Patent No. 6,177,078). In particular, US Patent No. 6,177,078 discloses the anti-IL-3Receptor alpha chain (IL-3Ra, CD123) monoclonal antibody 7G3, and the ability of 7G3 to bind to the N-terminal domain, specifically amino acid residues 19-49, of IL-3Ra. US Patent No. 6,733,743 discloses a method of impairing a hematologic cancer progenitor cell that expresses CD123 but does not significantly express CD131 , by contacting the cell with a composition of an antibody and a cytotoxic agent (selected from a chemotherapeutic agent, a toxin or an alpha-emitting radioisotope) whereby the composition binds selectively to CD123 in an amount effective to cause cell death. However, it has remained unclear whether targeting CD123 can functionally impair AML-LSCs.

The humanized anti-CD123 antibody "hz7G3" of which the sequences of the heavy and light chain variable domains are used in context of the "hz20G6Xhz7G3" antibody-like binding proteins comprises

a heavy chain variable domain consisting of sequence EVQLVQSGAEVKKPGESLKISCKGSGYSFTDYYMKWARQMPGKGLEWMGDIIPSSGAT FYNQKFKGQVTISADKSISTTYLQWSSLKASDTAMYYCARSHLLRASWFAYWGQGTMV TVSS (SEQ ID NO: 52, with CDRs shown in bold characters) comprising CDR1 -H of sequence SEQ ID NO: 50, a CDR2-H of sequence SEQ ID NO: 53, and a CDR3-H of sequence SEQ ID NO: 51 , and

- a light chain variable domain consisting of sequence DIVMTQSPDSLAVSLGERATINCESSQSLLNSGNQKNYLTWYQQKPGQPPKPLIYWAST RESGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQNDYSYPYTFGQGTKLEIK (SEQ ID NO: 54, with CDRs shown in bold characters) comprising CDR1 -L of sequence SEQ ID NO: 48, a CDR2-L of sequence WAS', and a CDR3-L of sequence SEQ ID NO: 49.

The humanized anti-CD123 antibody "hz7G3" comprises a N into S mutation at position 55 of SEQ ID NO: 52 in order to avoid the presence of a potential deamidation. As known to the skilled in the art, the presence of deamidation sites in antibodies are known to cause heterogeneity of antibody samples and thus preferably avoided.

Definitions

Throughout the instant application, the term "and/or" is a grammatical conjunction that is to be interpreted as encompassing that one or more of the cases it connects may occur. For example, the wording "such native sequence proteins can be prepared using standard recombinant and/or synthetic methods" indicates that native sequence proteins can be prepared using standard recombinant and synthetic methods or native sequence proteins can be prepared using standard recombinant methods or native sequence proteins can be prepared using synthetic methods.

Furthermore, throughout the instant application, the term "comprising" is to be interpreted as encompassing all specifically mentioned features as well optional, additional, unspecified ones. As used herein, the use of the term "comprising" also discloses the embodiment wherein no features other than the specifically mentioned features are present (i.e. "consisting of"). Furthermore the indefinite article "a^ or "an" does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

An "antibody" also called "immunoglobulin" may be a natural or conventional antibody in which two heavy chains are linked to each other by disulfide bonds and each heavy chain is linked to a light chain by a disulfide bond. There are two types of light chain, lambda (I) and kappa (k). There are five main heavy chain classes (or isotypes) which determine the functional activity of an antibody molecule: IgM, IgD, IgG, IgA and IgE. Each chain contains distinct sequence domains. The light chain includes two domains or regions, a variable domain (VL) and a constant domain (CL). The heavy chain includes four domains, a variable domain (VH) and three constant domains (CH 1 , CH2 and CH3, collectively referred to as CH). The variable regions of both light (VL) and heavy (VH) chains determine binding recognition and specificity to the antigen. The constant region domains of the light (CL) and heavy (CH) chains confer important biological properties such as antibody chain association, secretion, trans-placental mobility, complement binding, and binding to Fc receptors (FcR). The Fv fragment is the N-terminal part of the Fab fragment of an immunoglobulin and consists of the variable portions of one light chain and one heavy chain. The specificity of the antibody resides in the structural complementarity between the antibody combining site and the antigenic determinant. Antibody combining sites are made up of residues that are primarily from the hypervariable or complementarity determining regions (CDRs). Occasionally, residues

from nonhypervariable or framework regions (FR) influence the overall domain structure and hence the combining site. Complementarity Determining Regions or CDRs refer to amino acid sequences that together define the binding affinity and specificity of the natural Fv region of a native immunoglobulin binding site. The light and heavy chains of an immunoglobulin each have three CDRs, designated CDR1 -L, CDR2-L, CDR3-L and CDR1 -H, CDR2-H, CDR3-H, respectively. A conventional antibody antigen-binding site, therefore, includes six CDRs, comprising the CDR set from each of a heavy and a light chain V region.

In context of the invention, the antibody or immunoglobulin is an IgM, IgD, IgG, IgA and IgE.

"Framework Regions" (FRs) refer to amino acid sequences interposed between CDRs, i.e. to those portions of immunoglobulin light and heavy chain variable regions that are relatively conserved among different immunoglobulins in a single species. The light and heavy chains of an immunoglobulin each have four FRs, designated FR1 -L, FR2-L, FR3-L, FR4-L, and FR1 -H, FR2-H, FR3-H, FR4-H, respectively. Accordingly, the light chain variable domain may thus be designated as (FR1 -L)-(CDR1 -L)-(FR2-L)-(CDR2-L)-(FR3-L)-(CDR3-L)-(FR4-L) and the heavy chain variable domain may thus be designated as (FR1 -H)-(CDR1 -H)-(FR2-H)-(CDR2-H)-(FR3-H)-(CDR3-H)-(FR4-H).

Knowing the amino acid sequence of the CDRs one skilled in the art can easily determine the framework regions FR1 -L, FR2-L, FR3-L, FR4-L and/or FR1 -H, FR2-H, FR3-H, FR4-H.

As used herein, a "human framework region" is a framework region that is substantially identical (about 85%, or more, in particular 90%, 95%, 97%, 99% or 100%) to the framework region of a naturally occurring human antibody.

In the context of the invention, CDR/FR definition in an immunoglobulin light or heavy chain is to be determined based on IMGT definition (Lefranc et al. Dev. Comp.

Immunol., 2003, 27(1 ):55-77; www.imgt.org).

As used herein, the term "antibody" denotes conventional antibodies and fragments thereof, as well as single domain antibodies and fragments thereof, in particular variable heavy chain of single domain antibodies, and chimeric, humanized, bispecific or multispecific antibodies.

The term "humanized antibody" refers to an antibody which is wholly or partially of non-human origin and which has been modified to replace certain amino acids, in particular in the framework regions of the heavy and light chains, in order to avoid or minimize an immune response in humans. The constant domains of a humanized antibody are most of the time human CH and CL domains.

Numerous methods for humanization of an antibody sequence are known in the art; see e.g. the review by Almagro & Fransson (2008) Front Biosci. 13: 1619-1633. One commonly used method is CDR grafting, or antibody reshaping, which involves grafting of the CDR sequences of a donor antibody, generally a mouse antibody, into the framework scaffold of a human antibody of different specificity. Since CDR grafting may reduce the binding specificity and affinity, and thus the biological activity, of a CDR grafted non-human antibody, back mutations may be introduced at selected positions of the CDR grafted antibody in order to retain the binding specificity and affinity of the parent antibody. Identification of positions for possible back mutations can be performed using information available in the literature and in antibody databases. Amino acid residues that are candidates for back mutations are typically those that are located at the surface of an antibody molecule, while residues that are buried or that have a low degree of surface exposure will not normally be altered. An alternative humanization technique to CDR grafting and back mutation is resurfacing, in which non-surface exposed residues of non-human origin are retained, while surface residues are altered to human residues. Another alternative technique is known as "guided selection" (Jespers et al. (1994) Biotechnology 12, 899) and can be used to derive from for example a murine or rat antibody a fully human antibody conserving the epitope and binding characteristics of the parental antibody. A further method of humanization is the so-called 4D humanization. The 4D humanization protocol is described in the patent application US201 10027266 A1 (WO2009032661A1 ) and is exemplified in the following applying the 4D humanization to humanize the rat antibody variable light (VL) and heavy (VH) domains. In one example, a rat antibody homology model was done with typically MOE software (v. 201 1 .10-Chemical Computing Group, Quebec, Canada) using PDB structures (Berman et al., Nucleic Acids Research, 2000, 28:235-242) as templates and was subsequently energy minimized using the standard procedures implemented in MOE. A molecular dynamics (MD) simulation was then performed on the minimized 3D homology model (done with MOE software) of rat antibody and compared to the, for example, 49 human models derived from the seven representative light chains (vk1 , vk2, vk3, vk4, vlambdal , vlambda2, vlambda3) and the seven representative heavy chains (vhl a, vhl b, vh2, vh3, vh4, vh5, vh6) designed by LGCR/SDI and available within MOE. For instance, one model of chains couple (Vkx-Vhx) with the best both hydrophobic, electrostatic components and sequence identity outside CDR has been selected for the "4D humanization". For the pairwise association between the rat antibody variable domain and the selected model, the sequences were aligned based typically on the optimal 3D superposition of the alpha carbons of the corresponding homology models. The unwanted

motifs were then considered and mutated. Finally, the resulting humanized sequences were blasted for sequence similarity against, for instance, the IEDB database (http://www.immuneepitope.org; version 2012/01/30 accessible locally) to ensure that none of the sequences contain any known B- or T-cell epitope listed in.

For chimeric antibodies, humanization typically involves modification of the framework regions of the variable region sequences.

Amino acid residues that are part of a CDR will typically not be altered in connection with humanization, although in certain cases it may be desirable to alter individual CDR amino acid residues, for example to remove a glycosylation site, a deamidation site or an undesired cysteine residue. N-linked glycosylation occurs by attachment of an oligosaccharide chain to an asparagine residue in the tripeptide sequence Asn-X-Ser or Asn-X-Thr, where X may be any amino acid except Pro. Removal of an N-glycosylation site may be achieved by mutating either the Asn or the Ser/Thr residue to a different residue, in particular by way of conservative substitution. Deamidation of asparagine and glutamine residues can occur depending on factors such as pH and surface exposure. Asparagine residues are particularly susceptible to deamidation, primarily when present in the sequence Asn-Gly, and to a lesser extent in other dipeptide sequences such as Asn-Ala. When such a deamidation site, in particular Asn-Gly, is present in a CDR sequence, it may therefore be desirable to remove the site, typically by conservative substitution to remove one of the implicated residues. Substitution in a CDR sequence to remove one of the implicated residues is also intended to be encompassed by the present invention.

"Fragments" of (conventional) antibodies comprise a portion of an intact antibody, in particular the antigen binding region or variable region of the intact antibody. Examples of antibody fragments include Fv, Fab, F(ab')2, Fab', dsFv, (dsFv)2, scFv, sc(Fv)2, diabodies, bispecific and multispecific antibodies formed from antibody fragments. A fragment of a conventional antibody may also be a single domain antibody, such as a heavy chain antibody or VHH.

The term "Fab" denotes an antibody fragment having a molecular weight of about 50,000 and antigen binding activity, in which about a half of the N-terminal side of H chain and the entire L chain, among fragments obtained by treating IgG with a protease, papaine, are bound together through a disulfide bond.

The term "Fs domain" as used in context of the present invention encompasses native Fc and Fc variants and sequences as defined above. As with Fc variants and native Fc molecules, the term "Fc domain" includes molecules in monomeric or multimeric form, whether digested from whole antibody or produced by other means.

The term "native Fs" as used herein refers to a molecule comprising the sequence of a non-antigen-binding fragment resulting from digestion of an antibody or produced by other means, whether in monomeric or multimeric form, and can contain the hinge region. The original immunoglobulin source of the native Fc is, in particular, of human origin and can be any of the immunoglobulins, although IgGI and lgG2 are preferred. Native Fc molecules are made up of monomeric polypeptides that can be linked into dimeric or multimeric forms by covalent (i.e., disulfide bonds) and non-covalent association. The number of intermolecular disulfide bonds between monomeric subunits of native Fc molecules ranges from 1 to 4 depending on class (e.g., IgG, IgA, and IgE) or subclass (e.g., IgGI, lgG2, lgG3, IgAI, and lgGA2). One example of a native Fc is a disulfide-bonded dimer resulting from papain digestion of an IgG. The term "native Fc" as used herein is generic to the monomeric, dimeric, and multimeric forms.

The term "Fc variant" as used herein refers to a molecule or sequence that is modified from a native Fc but still comprises a binding site for the salvage receptor, FcRn (neonatal Fc receptor). Exemplary Fc variants, and their interaction with the salvage receptor, are known in the art. Thus, the term "Fc variant" can comprise a molecule or sequence that is humanized from a non-human native Fc. Furthermore, a native Fc comprises regions that can be removed because they provide structural features or biological activity that are not required for the antibody-like binding proteins of the invention. Thus, the term "Fc variant" comprises a molecule or sequence that lacks one or more native Fc sites or residues, or in which one or more Fc sites or residues has be modified, that affect or are involved in: (1 ) disulfide bond formation, (2) incompatibility with a selected host cell, (3) N-terminal heterogeneity upon expression in a selected host cell, (4) glycosylation, (5) interaction with complement, (6) binding to an Fc receptor other than a salvage receptor, or (7) antibody-dependent cellular cytotoxicity (ADCC).

The term "bispecific antibody" or "BsAb" typically denotes an antibody, which combines the antigen-binding sites of two antibodies within a single molecule. Thus, BsAbs are able to bind two different antigens simultaneously. Genetic engineering has been used with increasing frequency to design, modify, and produce antibodies or antibody derivatives with a desired set of binding properties and effector functions as described for instance in EP 2 050 764 A1.

The term "antibody-like binding protein" herein refers to polypeptides or binding proteins that, such as bispecific antibodies, are able to bind two different antigens simultaneously. However, different to conventional antibodies as defined herein antibodylike binding proteins comprise more than 6 CDRs. The antibody-like binding proteins of

the present invention are in the CODV format as defined herein below and are as further defined herein below in the section "Anti-CD3/anti-CD123 antibody-like binding proteins".

The "CODV format" in context of the present invention refers to the cross-over dual variable (CODV) configuration of bispecific antibodies or multispecific antibodies. The CODV format allows an interchangeability of variable domains with retention of folding and ultimate binding affinity.

The CODV format has been previously described in the international patent application WO2012/135345 and by Steinmetz et al. (MAbs. 2016 Jul; 8(5):867-78).

The term "linker" as used herein refers to one or more amino acid residues inserted between immunoglobulin domains to provide sufficient mobility for the domains of the light and heavy chains to fold into cross over dual variable region immunoglobulins. In some embodiments, a linker consists of 0 amino acid meaning that the linker is absent. A linker is inserted at the transition between variable domains or between variable and constant domains, respectively, at the sequence level. The transition between domains can be identified because the approximate size of the immunoglobulin domains is well understood. The precise location of a domain transition can be determined by locating peptide stretches that do not form secondary structural elements such as beta-sheets or alpha-helices as demonstrated by experimental data or as can be assumed by techniques of modeling or secondary structure prediction. The linkers described in context of the invention are the linkers L-i , L2, L3, L4 and L5. L-i is located between the N-terminal VD1 domain and the VD2 domain; L2 is located between the VD2 and the C-terminal CL domain. The linkers L3 and L4 are located on polypeptide as defined according to formula [III] of the antibody-like-proteins. More precisely, L3 is located between the N-terminal VD3 and the VD4 domains and L4 is located between the VD4 and the C-terminal CHi-Fc domains. L5 is located between CL and the N-terminal Fc2. The linkers L-i , L2, L3, L4 and L5 are independent, but in some embodiments, they have the same sequence and/or length. The linkers L-i , L2, L3, L4 and L5 are as defined herein above in context of the antibody-like binding proteins of the invention. Alternative linkers that might occur in variants of the antibody-like binding proteins of the invention are further described in the section "Variants of the anti-CD3/anti-CD123 antibody-like binding proteins".

The "RF mutation" generally refers to the mutation of the amino acids HY into RF in the CH3 domain of Fc domains, such as the mutation H435R and Y436F in CH3 domain as described by Jendeberg, L. et al. (1997, J. Immunological Meth., 201 : 25-34) and is described as advantageous for purification purposes as it abolishes binding to protein A.

In context of the present invention, the RF mutation refers for example to the position X6 and X7 of SEQ ID NO: 67, 68, 71 or 70, wherein the RF mutation is present when X6 is the amino acid R and X7 is the amino acid F. In one example, the RF mutation refers to the substitution of the amino acids HY with RF at positions 215-216 in Fc stump (Fc3) of SEQ ID NO: 69 (Fc3 of the antibody-like binding protein CODV-Fab-OL1 -Knobxhole-RF) or the mutation of HY into RF at positions 220-221 in the Fc region of sequence SEQ ID NO: 79 (Fc region of the antibody-like binding protein CODV-Fab-TL1 -Knobxhole-RF) as further described herein below in the section "Antibody-like-binding proteins".

The "Knob-into-Hole" or also called "Knob-into-Hole" technology refers to mutations Y349C, T366S, L368A and Y407V (Hole) and S354C and T366W (Knob) both in the CH3-CH3 interface to promote heteromultimer formation has been described in patents US5731 168 and US8216805, notably, which are herein incorporated by reference.

In context of the present invention, the "Knob" mutation refers for example to the position X2 and X3 of, for instance, SEQ ID NO: 66 or 62 wherein the Knob mutation is present when X2 is C and X3 is W. In one example, the Knob mutation refers to the substitutions S139C and T151W in the Fc of SEQ ID NO: 66 (Fc of the antibody-like binding protein CODV-Fab-OL1 a "hz20G6xhz7G3" and CODV-Fab-OL1 -Knobxhole-RF). In context of the present invention, the "hole" mutation refers for example to the position X-i , X3, X4 and X5 of, for instance, SEQ ID NO: 75 wherein the "hole" mutation is present when is C, X3 is S, X4 is A and X5 is V. In one example, the hole mutation refers to the substitutions Y134C, T151 S, L153A,Y192V in the Fc of SEQ ID NO: 75 (Fc of the antibodylike binding protein CODV-Fab-TL1 -Knob-RFxhole and CODV-Fab-TL1 -Knobxhole).

The "LALA mutation" refers to a double mutation L234A and L235A which abolishes Fc effector function. The Fc double mutant L234A and L235A does not bind FcyR or C1 q, and both ADCC and CDC functions of the Fc domain of lgG1 subclass are abolished (Hezareh, M. et al.,J Virol. 2001 Dec; 75(24): 12161-12168).

In context of the present invention, however, when referred to the double mutation

L234A and L235A the corresponding position may be different in the Fc domains as herein defined. However, the skilled in the art can easily identify the corresponding position in the Fc domain(s) (i.e. Fc in formula [III], Fc2 in formula [IV] and/or Fc3). In one example, the double mutation L234A and L235A corresponds to the double mutation L19A and L20A of Fc of sequence SEQ ID NO: 60, or in other words to mutation L359A and L358A in the polypeptide of formula [IV] of CODV-Fab-TL1 -RF of SEQ ID NO: 59.

By "purified" and "isolated" it is meant, when referring to a polypeptide (i.e. the antibody of the invention) or a nucleotide sequence, that the indicated molecule is present in the substantial absence of other biological macromolecules of the same type. The term "purified" as used herein in particular means at least 75%, 85%, 95%, or 98% by weight, of biological macromolecules of the same type are present. An "isolated" nucleic acid molecule that encodes a particular polypeptide refers to a nucleic acid molecule that is substantially free of other nucleic acid molecules that do not encode the subject polypeptide; however, the molecule may include some additional bases or moieties, which do not deleteriously affect the basic characteristics of the composition.

The term "antigen" or "target antigen" as used herein refers to a molecule or a portion of a molecule that is capable of being bound by an antibody or an antibody-like binding protein. The term further refers to a molecule or a portion of a molecule that is capable of being used in an animal to produce antibodies that are capable of binding to an epitope of that antigen. A target antigen may have one or more epitopes. With respect to each target antigen recognized by an antibody or by an antibody-like binding protein, the antibody-like binding protein is capable of competing with an intact antibody that recognizes the target antigen.

"Affinity" is defined, in theory, by the equilibrium association between the whole antibody and the antigen. Affinity may be expressed for example in half-maximal effective concentration (EC50) or the equilibrium dissociation constant (KD).

"Half maximal effective concentration" also called "EC50" refers to the concentration of a drug, antibody or toxicant which induces a response halfway between the baseline and maximum after a specified exposure time. EC50 and affinity are inversely related, the lower the EC50 value the higher the affinity of the antibody.

"KD" is the equilibrium dissociation constant, a ratio of koff/kon, between the antibody and its antigen. KD and affinity are inversely related. The KD value relates to the concentration of antibody and the lower the KD value and the higher the affinity of the antibody. Affinity can be experimentally assessed by a variety of known methods, such as measuring association and dissociation rates with surface Plasmon resonance or measuring the EC50 in an immunochemical assay (ELISA, flow cytometry). Enzyme-linked immunosorbent assay (ELISA) is a biochemistry assay that uses a solid-phase enzyme immunoassay to detect the presence of a substance, usually an antigen, in a liquid sample or wet sample. Antigens from the sample are attached to a surface. Then, a further specific antibody is applied over the surface so it can bind to the antigen. This antibody is linked to an enzyme, and, in the final step, a substance containing the

enzyme's substrate is added. The subsequent reaction produces a detectable signal, most commonly a color change in the substrate. Flow cytometry provides a method for analyzing a heterogeneous mixture of biological cells on single cell level based upon the specific light scattering and fluorescent characteristics or specific fluorescent labeling of each cell. In these assays, the EC50 is the concentration of the antibody which induces a response halfway between the baseline and maximum after some specified exposure time on a defined concentration of antigen by ELISA (enzyme-linked immuno-sorbent assay) or cells expressing the antigen by flow cytometry. Surface plasmon resonance is a label free method wherein the binding of a molecule in the soluble phase (the "analyte") is directly measured to a "Ngand" molecule immobilized on a sensor surface. In the sensor device the binding of the ligand is monitored by an optical phenomenon termed surface plasmon. In particular, when the "analyte" molecule dissociates from the "Ngand" molecule, a decrease in SPR signal (expressed in resonance units, RU) is observed. Association (On rate', ka) and Dissociation rates ('off rate', kd) are obtained from the signal obtained during the association and dissociation and the equilibrium dissociation constant ('binding constant', KD) can be calculated therefrom. The signal given in resonance units (RU) depends on the size of the ligand present in the analyte, however in case the experimental conditions are the same, i.e. the ligand is the same molecule at the same condition the obtained RU can indicate affinity, wherein the higher the obtained signal in RU the higher the binding.

A monoclonal antibody binding to antigen 1 (Ag1 ) is "cross-reactive" to antigen 2 (Ag2) when the EC50s are in a similar range for both antigens. In the present application, a monoclonal antibody binding to Ag1 is cross-reactive to Ag2 when the ratio of affinity of Ag2 to affinity of Ag1 is equal or less than 10 (in particular 5, 2, 1 or 0.5), affinities being measured with the same method for both antigens.

A monoclonal antibody binding to Ag1 is "not significantly cross-reactive" to Ag2 when the affinities are very different for the two antigens. Affinity for Ag2 may not be measurable if the binding response is too low. In the present application, a monoclonal antibody binding to Ag1 is not significantly cross-reactive to Ag2, when the binding response of the monoclonal antibody to Ag2 is less than 5% of the binding response of the same monoclonal antibody to Ag1 in the same experimental setting and at the same antibody concentration. In practice, the antibody concentration used can be the EC50 or the concentration required to reach the saturation plateau obtained with Ag1 .

As used herein "specificity" denotes the capacity of an antibody to discriminate the target peptide sequence to which its binds ("epitope") from closely related, highly homologous, peptide sequences.

A monoclonal antibody "binds specifically" to Ag1 when it is not significantly cross-reactive to Ag2.

The term "activation of T-cells" or "T-cell activation" herein refers to triggering CD3 signaling involving cytotoxic granule fusion, transient cytokine release, and proliferation. The antibody-like binding protein of the invention target CD3e and activate T-cells in the presence of target cells; this activity is also referred to as a "T-cell engaging effect". The T-cell engaging effect induces cytotoxicity in the target cell.

As known by the skilled in the art, activation of T-cells induces the expression of surface marker such as CD69 and CD25. The activation of T-cells can thus be measured by detecting and measuring the expression of CD4+/CD25+, CD4+/CD69+, CD8+/CD25+, or CD8+/CD69+ T cells. Methods to measure T-cell activation are known to the skilled in the art.

A method to measure T-cell activation is further disclosed in the example section (Example 2.9). Accordingly, in context of the invention T-cell activation is measured either as the percentage of cells expressing CD69 in % of the total number of cells, or as the percentage of cells expressing CD4 andCD69 in % of total number of cells, or as the percentage of cells expressing CD8 and CD69 in % of the total number of cells.

"Low T-cell activation" in the context of the antibody-like binding proteins of the invention refers to a T-cell activation less than 20%, less than 18%, less than 16%, less than 14%, less than 12%, less than 10%.

"Target cells" herein refer to cells that express the second antigen, in one example target cells herein refer to CD123 expressing cells such as THP-1 cells.

"High T-cell activation" herein refers to a T-cell activation higher than 50%, higher than 55%, higher than 60%, higher than 62%, higher than 64%, higher than 66%, higher than 68%, higher than 70%."Cvtotoxicitv" herein refers to the quality of a compound, such as the antibody-like binding protein of the invention, to be toxic to cells. Cytotoxicity may be induced by different mechanisms of action and can thus be divided into cell-mediated cytotoxicity, apoptosis, antibody-dependent cell-mediated cytotoxicity (ADCC) or complement-dependent cytotoxicity (CDC).

"Antibody-dependent cell-mediated cytotoxicity" or "ADCC" refers to a mechanism of cell-mediated immune defense whereby an effector cell of the immune system actively lyses a target cell, whose membrane-surface antigens have been bound by specific antibodies or antibody-like binding proteins of the invention.

"Complement-dependent cytotoxicity" or "CDC", in the context of the invention, refers to lysis of a target cell in the presence of complement system proteins.

"Cell-mediated cytotoxicity" refers to cytolysis of a target cell by effector lymphocytes, such as cytotoxic T lymphocytes or natural killer cells and can thus be distinguished into T-cell-mediated cytotoxicity and NK-cell cytotoxicity.

A "domain" may be any region of a protein, generally defined on the basis of sequence homologies and often related to a specific structural or functional entity.

A "recombinant" molecule is one that has been prepared, expressed, created, or isolated by recombinant means.

The term "gene" means a DNA sequence that codes for, or corresponds to, a particular sequence of amino acids which comprises all or part of one or more proteins or enzymes, and may or may not include regulatory DNA sequences, such as promoter sequences, which determine for example the conditions under which the gene is expressed. Some genes, which are not structural genes, may be transcribed from DNA to RNA, but are not translated into an amino acid sequence. Other genes may function as regulators of structural genes or as regulators of DNA transcription. In particular, the term gene may be intended for the genomic sequence encoding a protein, i.e. a sequence comprising regulator, promoter, intron and exon sequences.

A sequence "at least 85% identical to a reference sequence" is a sequence having, on its entire length, 85%, or more, in particular 90%, 91 %, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% sequence identity with the entire length of the reference sequence.

In the context of the present application, the "percentage of identity" is calculated using a global pairwise alignment (i.e. the two sequences are compared over their entire length). Methods for comparing the identity of two or more sequences are well known in the art. The « needle » program, which uses the Needleman-Wunsch global alignment algorithm (Needleman and Wunsch, 1970 J. Mol. Biol. 48:443-453) to find the optimum alignment (including gaps) of two sequences when considering their entire length, may for example be used. The needle program is for example available on the ebi.ac.uk World Wide Web site. The percentage of identity between two polypeptides, in accordance with the invention, is calculated using the EMBOSS: needle (global) program with a "Gap Open" parameter equal to 10.0, a "Gap Extend" parameter equal to 0.5, and a Blosum62 matrix.

Proteins consisting of an amino acid sequence "at least 80%, 85%, 90%, 95%, 96%, 97%, 98% or 99% identical" to a reference sequence may comprise mutations such as deletions, insertions and/or substitutions compared to the reference sequence. In case of substitutions, the protein consisting of an amino acid sequence at least 80%, 85%,

90%, 95%, 96%, 97%, 98% or 99% identical to a reference sequence may correspond to a homologous sequence derived from another species than the reference sequence.

"Amino acid substitutions" may be conservative or non-conservative. Preferably, substitutions are conservative substitutions, in which one amino acid is substituted for another amino acid with similar structural and/or chemical properties. The substitution preferably corresponds to a conservative substitution as indicated in the table below.

The terms "vector", "cloning vector" and "expression vector" mean the vehicle by which a DNA or RNA sequence (e.g. a foreign gene) can be introduced into a host cell, so as to transform the host and promote expression (e.g. transcription and translation) of the introduced sequence.

The term "transformation" means the introduction of a "foreign" (i.e. extrinsic) gene, DNA or RNA sequence to a host cell, so that the host cell will express the introduced gene or sequence to produce a desired substance, typically a protein or enzyme coded by the introduced gene or sequence. A host cell that receives and expresses introduced DNA or RNA bas been "transformed".

The term "expression system" means a host cell and compatible vector under suitable conditions, e.g. for the expression of a protein coded for by foreign DNA carried by the vector and introduced to the host cell.

The terms "pharmaceutical composition" or "therapeutic composition" as used herein refer to a compound or composition capable of inducing a desired therapeutic effect when properly administered to a patient.

Pharmaceutically" or "pharmaceutically acceptable" refers to molecular entities and compositions that do not produce an adverse, allergic or other untoward reaction when administered to a mammal, especially a human, as appropriate. A , "pharmaceutically-

acceptable carriers" or excipient refers to a non-toxic solid, semi-solid or liquid filler, diluent, encapsulating material or formulation auxiliary of any type.

As used herein, the term "subject" denotes a mammal, such as a rodent, a feline, a canine, and a primate. In particular, a subject according to the invention is a human.

The term "subject" or "individual" are used interchangeably and may be, for example, a human or a non-human mammal. For example, the subject is a bat; a ferret; a rabbit; a feline (cat); a canine (dog); a primate (monkey), an equine (horse); a human, including man, woman and child.

In the context of the invention, the term "treating" or "treatment", refers to a therapeutic use (i.e. on a subject having a given disease) and means reversing, alleviating, inhibiting the progress of one or more symptoms of such disorder or condition. Therefore, treatment does not only refer to a treatment that leads to a complete cure of the disease, but also to treatments that slow down the progression of the disease and/or prolong the survival of the subject.

By "preventing" is meant a prophylactic use (i.e. on a subject susceptible of developing a given disease).

The term "in need of treatment" refers to a subject having already the disorder as well as those in which the disorder is to be prevented.

By a "therapeutically effective amount" of the antibody-like binding protein or pharmaceutical composition thereof is meant a sufficient amount of the antibody-like binding protein to treat said cancer disease, at a reasonable benefit/risk ratio applicable to any medical treatment. It will be understood, however, that the total daily usage of the polypeptides and compositions of the present invention will be decided by the attending physician within the scope of sound medical judgment. The specific therapeutically effective dose level for any particular patient will depend upon a variety of factors including the disorder being treated and the severity of the disorder; activity of the specific polypeptide employed; the specific composition employed, the age, body weight, general health, sex and diet of the patient; the time of administration, route of administration, and rate of excretion of the specific polypeptide employed; the duration of the treatment; drugs used in combination or coincidental with the specific polypeptide employed; and like factors well known in the medical arts. For example, it is well known within the skill of the art to start doses of the compound at levels lower than those required to achieve the desired therapeutic effect and to gradually increase the dosage until the desired effect is achieved.

The "relapse" is defined as the reoccurrence of AML after complete remission.

"Complete remission" or "CR" is defined as follows: normal values for neutrophil (>1 .0*109/L), haemoglobin level of 10g/dl_ and platelet count (>100*109/L) and independence from red cell transfusion; blast cells less than 5%, no clusters or collections of blasts, and absence of Auer rods on bone marrow examination; and normal maturation of blood cells (morphology; myelogramme) and absence of extramedullary leukemia.

"Leukemic stem cells (LSCs)" are cancer cells that possess characteristics associated with normal stem cells, that is, the property of self renewal and the capability to develop multiple lineages. Such cells are proposed to persist in hematological cancers such as AML as distinct populations. The LCS present in AML patients are so called "AML-LCSs".

"Acute myelogenous leukemia (AML)" is a clonal disorder clinically presenting as increased proliferation of heterogeneous and undifferentiated myeloid blasts. The leukemic hierarchy is maintained by a small population of LSCs (AML-LCSs), which have the distinct ability for self-renewal, and are able to differentiate into leukemic progenitors. These progenitors generate the large numbers of leukemic blasts readily detectable in patients at diagnosis and relapse, leading ultimately to mortality. AML-LSC have been commonly reported as quiescent cells, in contrast to rapidly dividing clonogenic progenitors. This property of AML-LSCs renders conventional chemotherapeutics that target proliferating cells less effective, potentially explaining the current experience in which a high proportion of AML patients enter complete remission, but almost invariably relapse, with <30% of adults surviving for more than 4 years. In addition, minimal residual disease occurrence and poor survival has been attributed to high LSC frequency at diagnosis in AML patients. Consequently, it is imperative for the long-term management of AML (and similarly other above mentioned hematological cancer conditions) that new treatments are developed to specifically eliminate LSCs. Over-expression of CD123 has been reported on AML blasts and on CD34+/CD38 AML- LSCs relative to normal hematopoietic cells.

Anti-CD3/anti-CD123 antibody-like binding proteins

For purposes of simplicity, throughout the instant application, "anti-CD3/anti-CD123 antibody-like binding proteins" or "anti-CD3/anti-CD123 antibody-like binding proteins of the invention" might be referred to as "antibody-like binding proteins" or "antibody-like binding proteins of the invention.

These antibody-like binding proteins have a CODV design.

Accordingly, in one embodiment, the antibody-like binding protein of the invention is in the CODV format as previously described in the international patent application WO2012/135345, which is incorporated herein by reference.

In one embodiment, the antibody-like binding protein of the invention is in the CODV format as previously described in the international patent application WO2012/135345, wherein the light chain is elongated with an additional Fc domain. Each light chain and heavy chain comprise a Fc domain. Those antibody-like binding proteins of the invention are CODV-Fab-TL1 "hz20G6Xhz7G3" antibody-like binding proteins.

In one embodiment, the antibody-like binding protein of the invention is in the CODV format as previously described in the international patent application WO2012/135345, wherein there is an additional Fc domain. The heavy chain comprises a Fc domain, but not the light chain. Those antibody-like binding proteins of the invention are CODV-Fab-OL1 "hz20G6Xhz7G3" antibody-like binding proteins.

In one embodiment, the invention refers to an antibody-like binding protein that binds specifically to human CD3e and human CD123 comprising two polypeptide chains that form two antigen-binding sites, wherein one polypeptide chain has a structure represented by the formula [I]:

VD1-L VD2-L2-CL [I] and one polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHI-Fc [III]

wherein:

a) one polypeptide of formula [I] consists of the amino acid sequence SEQ ID NO: 55 which comprises VD1 of sequence SEQ ID NO: 54, U of sequence SEQ ID NO: 56, VD2 of sequence SEQ ID NO: 10, L2 of sequence SEQ ID NO: 56, CL of sequence SEQ ID NO: 18, or

a sequence at least 85% identical to SEQ ID NO: 55 in which the 3 CDRs of sequences SEQ ID NO: 48, AS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered; and

b) one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO : 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, CHi of sequence SEQ ID NO: 19, and Fc of sequence SEQ ID NO: 68 wherein X^ is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R and X7 is Y or F, or

a sequence at least 85% identical to SEQ ID NO: 67 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 are unaltered, and said amino acids Xi, X2, X3, X4, X5, Χβ and X7 are as defined above,

and wherein the polypeptide formula [I] and the polypeptide of formula [III] form a crossover light chain-heavy chain pair.

In one embodiment, the antibody-like binding protein as defined herein above does not comprise an antibody-like binding protein wherein the polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, Cm of sequence SEQ ID NO: 19, and Fc of sequence SEQ ID NO: 68 wherein X^ is Y, X2 is S, X3 is T, X4 is L, X5 is Y, X6 is H and X7 is Y, and/or the antibody-like binding protein as defined herein above does not comprise an antibody-like binding protein wherein the polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, Cm of sequence SEQ ID NO: 19, and Fc of sequence SEQ I D NO: 68 wherein X-\ is Y, X2 is C, X3 is W, X4 is L, X5 is Y, and X6 is H and X7 is Y, or X6 is R and X7 is F.

Accordingly, in one embodiment, the antibody-like binding protein as defined herein above does not comprise an antibody-like binding protein wherein the polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 59, and/or

the antibody-like binding protein as defined herein above does not comprise an antibody-like binding protein wherein the polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 61 or SEQ ID NO: 65.

In one embodiment, the antibody-like binding protein of the present invention does not comprise:

a) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID

NO: 57; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 59, and/or

b) one polypeptide of formula [I] consisting of the amino acid sequence SEQ ID NO: 55; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 61 , and the polypeptide Fc stump (Fc3) of SEQ ID NO: 63; and/or

c) one polypeptide of formula [I] consisting of the amino acid sequence SEQ ID NO: 55; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO:

65, and the polypeptide Fc stump (Fc3) of SEQ ID NO: 64.

The antibody-like binding proteins are so called "hz20G6Xhz7G3" antibody-like binding proteins because polypeptide [I] comprises VD1 and VD2 that are the variable domains of the light chains of the humanized anti-CD123 antibody "7G3" (also called "hz7G3") and humanized anti-CD3 antibody "20G6" (also called "hz20G6"), respectively, and polypeptide [III] comprises VD3 and VD4 that are the variable domains of the heavy chains of the humanized anti-CD3 antibody "20G6" (also called "hz20G6) and humanized anti-CD123 antibody "7G3" (also called "hz7G3"), respectively.

More particularly, the antibody-like binding proteins are so called "hz20G6Xhz7G3" antibody-like binding proteins because the polypeptide chain having a structure represented by the formula [I] comprises VD1 of sequence SEQ ID NO: 54 which is the light chain variable domain of the humanized anti-CD123 antibody "7G3" (also called "hz7G3") and VD2 of sequence SEQ ID NO: 10 which is the light chain variable domain amino acid sequence VL1 c of humanized anti-CD3 antibody "20G6", and the polypeptide chain having a structure represented by the formula [III] comprises VD3 of sequence SEQ ID NO: 9 which is the heavy chain variable domain variant VH1 d of humanized anti-CD3 antibody "20G6" and VD4 of sequence SEQ ID NO: 52 which is a variant heavy chain variable domain of the humanized anti-CD123 antibody "7G3" (also called "hz7G3").

As defined above, the polypeptide chain having a structure represented by the formula [III] comprises the Fc of sequence SEQ ID NO: 68 wherein X-i is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R and X7 is Y or F.

It will be understood by the skilled in the art, that a Fc sequence of SEQ ID NO: 68

- wherein is Y, X2 is S, X3 is T, X4 is L, X5 is Y and X6 is H and X7 is Y is a so- called wild-type Fc sequence of SEQ ID NO: 60, and

- wherein is Y, X2 is S, X3 is T, X4 is L, X5 is Y and X6 is R and X7 is F is a Fc sequence comprising the RF mutation, and

- wherein is C, X2 is S, X3 is S, X4 is A, X5 is V and X6 is H and X7 is Y is a Fc sequence comprising the hole mutation as defined herein above in the section "definitions" and results in a Fc domain of SEQ ID NO: 75, and

- wherein X-i is C, X2 is S, X3 is S, X4 is A, X5 is V and X6 is R and X7 is F is a Fc sequence comprising the hole mutation and RF mutation as already defined herein above and results in a Fc domain of SEQ ID NO: 79, and

- wherein X-i is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is H and X7 is Y is a Fc sequence comprising the Knob mutation as already defined herein above in the section "definitions" and results in a Fc domain of SEQ ID NO: 66, and

- wherein X-i is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is R and X7 is F is a Fc sequence comprising the Knob mutation and RF mutation as already defined herein above in the section definitions" and results in a Fc domain of SEQ ID NO: 62.

It will be further understood that the general definition according to which X-i is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R and X7 is Y or F in any of the Fc domains, i.e Fc of SEQ ID NO: 68 and FC and Fc2 domain of SEQ ID NO: 70 (Fc2 domain of SEQ ID NO: 70 is introduced herein below), may be replaced in all embodiments wherein is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R and X7 is Y or F with the definition according to which,

- X-\ is Y, X2 is S, X3 is T, X4 is L, X5 is Y (corresponding to wild-type), or

- Xi is C, X2 is S, X3 is S, X4 is A, X5 is V (corresponding to "hole" mutation), or

- X-i is Y, X2 is C, X3 is W, X4 is L, X5 is Y (corresponding to "Knob" mutation), and

- X6 is H and X7 is Y (corresponding to wild-type), or

- X6 is R and X7 is F (corresponding to "RF" mutation).

Accordingly, for further exemplification, in one embodiment, the invention refers to an antibody-like binding protein that binds specifically to human CD3e and human CD123 comprising two polypeptide chains that form two antigen-binding sites, wherein one polypeptide chain has a structure represented by the formula [I]:

VD1-L VD2-L2-CL [I] and one polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHI-Fc [III] wherein:

a) one polypeptide of formula [I] consists of the amino acid sequence SEQ ID NO: 55 which comprises VD1 of sequence SEQ ID NO: 54, U of sequence SEQ ID NO: 56, VD2 of

sequence SEQ ID NO: 10, L2 of sequence SEQ ID NO: 56, CL of sequence SEQ ID NO: 18, or

a sequence at least 85% identical to SEQ ID NO: 55 in which the 3 CDRs of sequences SEQ ID NO: 48, AS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered; and

b) one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, Cm of sequence SEQ ID NO: 19, and Fc of sequence SEQ ID NO: 68 wherein

Xi is Y, X2 is S, X3 is T, X4 is L, X5 is Y, or

Xi is C, X2 is S, X3 is S, X4 is A, X5 is V, or

Xi is Y, X2 is C, X3 is W, X4 is L, X5 is Y, and

X6 is H and X7 is Y, or

X6 is R and X7 is F, or

a sequence at least 85% identical to SEQ ID NO: 67 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 are unaltered, and said amino acids Xi, X2, X3, X4, X5, Χβ and X7 are as defined above,

and wherein the polypeptide formula [I] and the polypeptide of formula [III] form a crossover light chain-heavy chain pair.

In a further embodiment, a second Fc domain (called Fc2) is added to the polypeptide of formula [I] of the antibody-like binding protein CODV-Fab.

Accordingly, in one embodiment, the polypeptide of formula [I] further comprises a Fc domain (Fc2). In the same embodiment a linker L5 is present between CL and the Fc2 domain of the polypeptide chains of formula [I] resulting in the polypeptide chains of formula [IV].

In one particular embodiment, the polypeptide of formula [I] further comprises the Fc2 domain of SEQ ID NO: 70 , wherein Xi is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R and X7 is Y or F.

Accordingly, the invention further refers to an antibody-like binding protein comprising two polypeptide chains that form two antigen-binding sites, wherein one polypeptide chain has a structure represented by the formula [IV]:

VD1-L VD2-L2-CL-L5-Fc2 [IV] and one polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHI-Fc [III] wherein:

a) one polypeptide of formula [IV] consists of the amino acid sequence SEQ ID NO: 71 which comprises VD1, U, VD2, L2 and CL as defined above for the polypeptide chain represented by the formula [I] and L5 which consists of 0 amino acid and Fc2 of sequence SEQ ID NO: 70 wherein Xi is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R and X7 is Y or F, or

a sequence at least 85% identical to SEQ ID NO: 71 in which the 3 CDRs of sequences SEQ ID NO: 48, AS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered and said amino acids Xi, X2, X3, X4, X5, X6 and X7 in SEQ ID NO: 71 in said polypeptide chain represented by the formula [IV] are as defined above;

b) one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, CHi of sequence SEQ ID NO: 19, and Fc of sequence SEQ ID NO: 68 wherein Xi is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R and X7 is Y or F, or

a sequence at least 85% identical to SEQ ID NO: 67 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO:53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 are unaltered, and said amino acids Xi, X2, X3, X4, X5, Χβ and X7 in SEQ ID NO: 67 are as defined above, and

wherein the polypeptide formula [IV] and the polypeptide of formula [III] form a cross-over light chain-heavy chain pair.

This CODV format, in which the polypeptide chains represented by the formula [III] and [IV] dimerizes through their respective Fc2 and Fc regions, is herein called CODV-Fab-TL.

In a related embodiment, the antibody-like binding protein as defined herein above does not comprise an antibody-like binding protein wherein

a) the polypeptide of formula [IV] consists of the amino acid sequence SEQ ID NO: 71 which comprises VD1, U, VD2, L2 and CL as defined above for the polypeptide chain represented by the formula [I] and L5 which consists of 0 amino acid and Fc2 of sequence SEQ ID NO: 70 wherein X^ is Y, X2 is S, X3 is T, X4 is L, X5 is Y, X6 is

R and X7 is F, and

b) the polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO : 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, Cm of sequence SEQ ID NO: 19, and FC of sequence SEQ ID NO: 68 wherein X^ is Y, X2 is S, X3 is T, X4 is L, X5 is Y, X6 is H and X7 is Y.

Accordingly, in one embodiment, the antibody-like binding protein as defined herein above does not comprise an antibody-like binding protein wherein

a) the polypeptide of formula [IV] consists of the amino acid sequence SEQ ID NO: 57, and

b) the polypeptide of formula [III] consists of the amino acid sequence SEQ ID

NO : 59.

In a further related embodiment, the invention refers to an antibody-like binding protein that binds specifically to human CD3e and human CD123 comprising two polypeptide chains that form two antigen-binding sites, wherein one polypeptide chain has a structure represented by the formula [IV]:

Vm-L ViM-Lz-CL-U-Fcz [IV] and one polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHI-Fc [III] wherein:

a) said polypeptide of formula [IV] consists of:

(i) the amino acid sequence SEQ ID NO: 71 which comprises

VDi of sequence SEQ ID NO: 54,

U of sequence SEQ ID NO: 56,

· VD2 of sequence SEQ ID NO: 10,

L2 of sequence SEQ ID NO: 56,

CL of sequence SEQ ID NO: 18,

L5 consists of 0 amino acid, and

♦ FC2 consists of sequence SEQ ID NO: 70

· wherein X-\ is Y, X2 is S, X3 is T, X4 is L, X5 is Y and X6 is H and X7 is Y, or

• wherein X^ is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is H and X7 is Y, or

• wherein X^ is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is R and X7 is F,

or

(ii) a sequence at least 85% identical to SEQ ID NO: 71 in which

• the 3 CDRs of sequences SEQ ID NO: 48, 'WAS' and SEQ ID NO:

49 of VD1 of sequence SEQ ID NO: 54 are unaltered, and

• the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered, and

• the amino acids Xi, X2, X3, X4, Xs, Xe and X7 in SEQ ID NO: 71 are as defined above in a)(i);

b) said polypeptide of formula [III] consists of:

(i) the amino acid sequence SEQ ID NO : 67 which comprises

• VD3 of sequence SEQ ID NO: 9,

l_3 which consists of 0 amino acid,

VD4 of sequence SEQ ID NO: 52,

L4 which consists of 0 amino acid,

Cm of sequence SEQ ID NO: 19, and

• FC consists of sequence SEQ ID NO: 68, wherein is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R, and X7 is Y or F,

or

(ii) a sequence at least 85% identical to SEQ ID NO: 67 in which

• the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52 are unaltered, and

• the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 are unaltered, and

• the amino acids Xi, X2, X3, X4, X5, Xe and X7 of SEQ ID NO: 67 are as defined above in b)(i),

and wherein the polypeptide formula [IV] and the polypeptide of formula [III] form a crossover light chain-heavy chain pair.

In a further related embodiment, the invention refers to an antibody-like binding protein comprising two polypeptide chains that form two antigen-binding sites, wherein one polypeptide chain has a structure represented by the formula [IV]:

VD1-L VD2-L2-CL-L5-FC2 [IV] and one polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHI-Fc [III] wherein:

a) one polypeptide of formula [IV] consists of the amino acid sequence SEQ ID NO: 71 which comprises VD1, U, VD2, L2 and CL as defined above and L5 consists of 0 amino acid and Fc2 of sequence SEQ ID NO: 70 wherein

wherein Xi is Y, X2 is S, X3 is T, X4 is L, X5 is Y and X6 is H and X7 is Y, or

wherein Xi is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is H and X7 is Y, or

wherein Xi is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is R and X7 is F, or

a sequence at least 85% identical to SEQ ID NO: 71 in which the 3 CDRs of sequences SEQ ID NO: 48, AS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered and said amino acids Xi , X2, X3, X4, X5, Xe and X7 in SEQ ID NO: 71 are as defined above;

b) one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO : 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, Cm of sequence SEQ ID NO: 19, and Fc of sequence SEQ ID NO: 68 wherein Xi is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R and X7 is Y or F, or

a sequence at least 85% identical to SEQ ID NO: 67 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 are unaltered, and said amino acids Xi , X2, X3, X4, X5, Xe and X7 of SEQ ID NO: 67 are as defined above

and wherein the polypeptide formula [IV] and the polypeptide of formula [III] form a crossover light chain-heavy chain pair.

It will be understood by the skilled in the art, that when one Fc domain is wild-type sequence or carries the Knob mutations the other Fc domain is either wild-type or carries the hole mutation.

Accordingly, in one further related embodiment, the antibody-like binding protein according to the invention comprises

a) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

71 which comprises VD1, U, VD2, L2 and CL as defined above and L5 consists of 0 amino acid and Fc2 of sequence SEQ ID NO: 70 wherein

wherein is Y, X2 is S, X3 is T, X4 is L, X5 is Y and X6 is H and X7 is Y, or

wherein Xi is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is H and X7 is Y, or

wherein Xi is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is R and X7 is F, and

b) one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO : 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, Cm of sequence SEQ ID NO: 19, and Fc of sequence SEQ ID NO: 68 wherein

Xi is Y, X2 is S, X3 is T, X4 is L, X5 is Y, or

Xi is C, X2 is S, X3 is S, X4 is A, X5 is V, and

X6 is H and X7 is Y, or

X6 is R and X7 is F.

Accordingly, in one particular embodiment, the invention further refers to an antibody-like binding protein comprising two polypeptide chains that form two antigen-binding sites, wherein one polypeptide chain has a structure represented by the formula [IV]:

VD1-L VD2-L2-CL-L5-FC2 [IV] and one polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHI-Fc [III] wherein:

a) one polypeptide of formula [IV] consists of the amino acid sequence SEQ ID NO: 71 which comprises VD1, U, VD2, L2 and CL as defined above for the polypeptide chain represented by the formula [I] and L5 which consists of 0 amino acid and FC2 of sequence SEQ ID NO: 70 wherein

Xi is Y, X2 is C, X3 is W, X4 is L, X5 is Y, and X6 is H and X7 is Y, or X6 is R and X7 is F, or a sequence at least 85% identical to SEQ ID NO: 71 in which the 3 CDRs of sequences SEQ ID NO: 48, AS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered and said amino acids Xi, X2, X3, X4, X5, X6 and X7 in SEQ ID NO: 70 in said polypeptide chain represented by the formula [IV] are as defined above;

b) one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, CH1 of sequence SEQ ID NO: 19, and FC of sequence SEQ ID NO: 68 wherein

Xi is C, X2 is S, X3 is S, X4 is A, X5 is V, and X6 is H and X7 is Y, or X6 is R and X7 is F, or a sequence at least 85% identical to SEQ ID NO: 67 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO:53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of

sequence SEQ ID NO: 9 are unaltered, and said amino acids Xi, X2, X3, X4, X5, Χβ and X7 in SEQ ID NO: 67 are as defined above,

and wherein the polypeptide formula [IV] and the polypeptide of formula [III] form a crossover light chain-heavy chain pair.

Accordingly, in one embodiment, the antibody-like binding protein comprises:

a) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 81 or a sequence at least 85% identical to SEQ ID NO: 81 , and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 60 or a sequence at least 85% identical to SEQ ID NO: 60, or

b) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 73 or a sequence at least 85% identical to SEQ ID NO: 73, and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 75 or a sequence at least 85% identical to SEQ ID NO: 75, or

c) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 77 or a sequence at least 85% identical to SEQ ID NO: 77, and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 75 or a sequence at least 85% identical to SEQ ID NO: 75, or

d) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 77 or a sequence at least 85% identical to SEQ ID NO: 77, and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 79 or a sequence at least 85% identical to SEQ ID NO: 79.

Accordingly, in one further embodiment, the antibody-like binding protein comprises: i) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 73 or a sequence at least 85% identical to SEQ ID NO: 73, and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 75 or a sequence at least 85% identical to SEQ ID NO: 75, or

ii) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 77 or a sequence at least 85% identical to SEQ ID NO: 77, and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 79 or a sequence at least 85% identical to SEQ ID NO: 79.

In a further embodiment, the antibody-like binding protein according to the invention is selected from the group consisting of antibody-like binding proteins wherein: a) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 81 , and the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 60, or

b) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 73, and the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 75, or

c) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 77, and the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 75, or

d) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 77, and the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 79.

In a further embodiment, the antibody-like binding protein according to the invention is selected from the group consisting of antibody-like binding proteins wherein:

i) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 73, and the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 75, or

ii) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 77, and the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 79.

In a further embodiment, the antibody-like binding molecule comprises:

a) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

80, or

a sequence at least 85% identical to SEQ ID NO: 80 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 of SEQ ID NO: 80 are unaltered; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 59, or

a sequence at least 85% identical to SEQ ID NO: 59 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 , 559, 560, 478, 490 of SEQ ID NO: 59 are unaltered; or

b) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

72, or

a sequence at least 85% identical to SEQ ID NO: 72 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 of SEQ ID NO: 72 are unaltered; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 74, or

a sequence at least 85% identical to SEQ ID NO: 74 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 , 559, 560, 478, 490 of SEQ ID NO: 74 are unaltered; or

one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO: 76, or

a sequence at least 85% identical to SEQ ID NO: 76 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 in SEQ ID NO: 76 are unaltered; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 74, or

a sequence at least 85% identical to SEQ ID NO: 74 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 , 559, 560, 478, 490 of SEQ ID NO: 74 are unaltered; or

one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO: 76, or

a sequence at least 85% identical to SEQ ID NO: 76 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 in SEQ ID NO: 76 are unaltered; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 78, or

a sequence at least 85% identical to SEQ ID NO: 78 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 , 559, 560, 478, 490 of SEQ ID NO: 78 are unaltered.

a further embodiment, the antibody-like binding molecule comprises:

one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO: 72, or

a sequence at least 85% identical to SEQ ID NO: 72 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 of SEQ ID NO: 72 are unaltered; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 74, or

a sequence at least 85% identical to SEQ ID NO: 74 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 , 559, 560, 478, 490 of SEQ ID NO: 74 are unaltered; or

one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO: 76, or

a sequence at least 85% identical to SEQ ID NO: 76 in which the 3 CDRs of sequences SEQ I D NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 in SEQ ID NO: 76 are unaltered; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 78, or

a sequence at least 85% identical to SEQ ID NO: 78 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 , 559, 560, 478, 490 of SEQ ID NO: 78 are unaltered.

In a further embodiment, the antibody-like binding protein comprises:

i) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

80; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 59,

ii) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

72, and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO:

74,

iii) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

76; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 74, and

iv) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

76; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 78.

In a further embodiment, the antibody-like binding protein comprises:

a) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

72, and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 74,

b) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

76; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 78.

In one embodiment, the antibody-like binding protein comprising one polypeptide chain having a structure represented by the formula [I] and one polypeptide chain having a structure represented by the formula [III] as defined herein above, further comprises a third polypeptide chain comprising a Fc domain (called Fc3).

It will be understood by the skilled in the art that said FC3 domain might be referred to as a second Fc domain, because second polypeptide having a structure represented by the formula [III] comprises a first Fc domain.

Accordingly, in one embodiment, the invention refers to an antibody-like binding protein that binds specifically to human CD3e and human CD123 comprising three polypeptide chains that form two antigen-binding sites, wherein

a first polypeptide has a structure represented by the formula [I]:

VD1-L VD2-L2-CL [I] and a second polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHI-Fc [III];

and a third polypeptide Fc3 which is the immunoglobulin hinge region and CH2, CH3 immunoglobulin heavy chain constant domains of an immunoglobulin;

wherein

a) said polypeptide of formula [I] consists of:

(i) the amino acid sequence SEQ ID NO: 55 which comprises

· VD1 of sequence SEQ ID NO: 54,

• Li of sequence SEQ ID NO: 56,

• VD2 of sequence SEQ ID NO: 10,

• L2 of sequence SEQ ID NO: 56,

• CL of sequence SEQ ID NO: 18,

or

(ii) a sequence at least 85% identical to SEQ ID NO: 55 in which

• the 3 CDRs of sequences SEQ ID NO: 48, 'WAS' and SEQ ID NO:

49 of VD1 of sequence SEQ ID NO: 54, are unaltered and

• the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered;

b) said polypeptide of formula [III] consists of:

(i) the amino acid sequence SEQ ID NO: 67 which comprises:

• VD3 of sequence SEQ ID NO: 9,

• l_3 which consists of 0 amino acid,

· VD4 of sequence SEQ ID NO: 52,

• l_4 which consists of 0 amino acid,

• CH1 of sequence SEQ ID NO: 19, and

• Fc of sequence SEQ ID NO: 68 wherein Xi is Y, X2 is C, X3 is W, X4 is L, X5 is Y, and X6 is H and X7 is Y, or X6 is R and X7 is F, or

(ii) a sequence at least 85% identical to SEQ ID NO : 67 in which

• the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO:53, and SEQ

ID NO: 51 of VD4 of sequence SEQ ID NO: 52, are unaltered and

• the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 are unaltered, and

• the amino acids Xi, X2, X3, X4, X5, Χβ and X7 are as defined above in b)(i),

and wherein:

- the polypeptide formula [I] and the polypeptide of formula [III] form a cross-over light chain-heavy chain pair,

- the polypeptide of formula [III] heterodimerizes with the third polypeptide through its Fc domain

- said third polypeptide Fc3 consists of SEQ ID NO: 69 or a sequence at least 85% identical to SEQ ID NO: 69, wherein the amino acid positions 129, 146, 148, 187, 215, 216 of SEQ ID NO: 69 are unaltered.

Accordingly, in one embodiment, the invention refers to an antibody-like binding protein which comprises three polypeptide chains that form two antigen-binding sites, wherein

a first polypeptide has a structure represented by the formula [I]:

VD1-L VD2-L2-CL [I] and a second polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHi-Fc [III];

and a third polypeptide Fc3 (also called "Fc stump") which is the immunoglobulin hinge region and CH2, CH3 immunoglobulin heavy chain constant domains of an immunoglobulin; wherein

a) one polypeptide of formula [I] consists of the amino acid sequence SEQ ID NO: 55 which comprises VD1 of sequence SEQ ID NO: 54, U of sequence SEQ ID NO: 56, VD2 of sequence SEQ ID NO: 10, L2 of sequence SEQ ID NO: 56, CL of sequence SEQ ID NO: 18, or

a sequence at least 85% identical to SEQ ID NO: 55 in which the 3 CDRs of sequences SEQ ID NO: 48, AS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered;

b) one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO : 67 which comprises VD3 of sequence SEQ ID NO: 9, L3 which consists of 0 amino acid, VD4 of sequence SEQ ID NO: 52, L4 which consists of 0 amino acid, CH1 of sequence SEQ ID NO: 19, and Fc of sequence SEQ ID NO: 68 wherein

Xi is Y, X2 is C, X3 is W, X4 is L, X5 is Y, and X6 is H and X7 is Y, or X6 is R and X7 is F, or a sequence at least 85% identical to SEQ ID NO : 67 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO:53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52,

and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 are unaltered, and said amino acids Xi, X2, X3, X4, X5, Χβ and X7 are as defined above

and wherein the polypeptide formula [I] and the polypeptide of formula [III] form a cross-over light chain-heavy chain pair,

and wherein the polypeptide of formula [III] heterodimerizes with the third polypeptide through its Fc domain.

Accordingly, in said embodiment, the so-called "Fc stump" (Fc3) heterodimerizes with the Fc region of the polypeptide of formula [III]. This CODV format is herein called CODV-Fab-OL. This construct avoids that the CODV-Fab form aggregates.

Accordingly, in one particular embodiment, the Fc3 domain of the antibody-like binding protein as defined above consists of SEQ ID NO: 69.

In a related embodiment, the antibody-like binding protein according to the invention comprises

- the polypeptide of formula [I] consisting of SEQ ID NO: 55, or

a sequence at least 85% identical to SEQ ID NO: 55 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered; and

the polypeptide of formula [III] comprising the Fc domain of sequence SEQ ID NO: 66, or

a sequence at least 85% identical to SEQ ID NO: 59 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ I D NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 , 539, 560, 478, 490 of SEQ ID NO: 66 are unaltered; and

- Fc stump (Fc3) consisting of SEQ ID NO: 69 or a sequence at least 85% identical to SEQ ID NO: 69, wherein the the amino acid positions 129, 146, 148, 187, 215,

216 of SEQ ID NO: 69 are unaltered.

CLAIMS
1 . An antibody-like binding protein that binds specifically to human CD3e and human CD123 comprising two polypeptide chains that form two antigen-binding sites, wherein one polypeptide chain has a structure represented by the formula [IV]:

Vo L VorLz-CL-Ls-Fcz [IV] and one polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHI-Fc [III] wherein:

b) said polypeptide of formula [IV] consists of:

(i) the amino acid sequence SEQ ID NO: 71 which comprises

• VDi of sequence SEQ ID NO: 54,

• L-i of sequence SEQ ID NO: 56,

• VD2 of sequence SEQ ID NO: 10,

· l_2 of sequence SEQ ID NO: 56,

• CL of sequence SEQ ID NO: 18,

• L5 consists of 0 amino acid, and

• FC2 consists of sequence SEQ ID NO: 70

• wherein X-\ is Y, X2 is S, X3 is T, X4 is L, X5 is Y and X6 is H and X7 is Y, or

• wherein X^ is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is H and X7 is Y, or

• wherein X-\ is Y, X2 is C, X3 is W, X4 is L, X5 is Y and X6 is R and X7 is F,

or

(iii) a sequence at least 85% identical to SEQ ID NO: 71 in which

• the 3 CDRs of sequences SEQ ID NO: 48, 'WAS' and SEQ ID NO:

49 of VD1 of sequence SEQ ID NO: 54 are unaltered, and

• the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered, and

• the amino acids Xi, X2, X3, X4, X5, Xe and X7 in SEQ ID NO: 71 are as defined above in a)(i);

b) said polypeptide of formula [III] consists of:

(i) the amino acid sequence SEQ ID NO : 67 which comprises

· VD3 of sequence SEQ ID NO: 9,

• l_3 which consists of 0 amino acid,

• VD4 of sequence SEQ ID NO: 52,

• L4 which consists of 0 amino acid,

• Cm of sequence SEQ ID NO: 19, and

• Fc consists of sequence SEQ ID NO: 68, wherein is Y or C, X2 is S or C, X3 is T, S or W, X4 is A or L, X5 is V or Y, X6 is H or R, and X7 is Y or F,

or

(iii) a sequence at least 85% identical to SEQ ID NO: 67 in which

• the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52 are unaltered, and

• the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 are unaltered, and

• the amino acids Xi, X2, X3, X4, X5, Xe and X7 of SEQ ID NO: 67 are as defined above in b)(i),

and wherein the polypeptide formula [IV] and the polypeptide of formula [III] form a crossover light chain-heavy chain pair.

2. The antibody-like binding protein according to claim 1 , wherein the polypeptide of formula [III] comprises the Fc of sequence SEQ ID NO: 68 wherein

Xi is Y, X2 is S, X3 is T, X4 is L, X5 is Y, or

Xi is C, X2 is S, X3 is S, X4 is A, X5 is V, and

X6 is H and X7 is Y, or

X6 is R and X7 is F.

3. The antibody-like binding protein according to claim 1 or 2, wherein

a) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 81 or a sequence at least 85% identical to SEQ ID NO: 81 , and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 60 or a sequence at least 85% identical to SEQ ID NO: 60, or

b) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 73 or a sequence at least 85% identical to SEQ ID NO: 73, and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 75 or a sequence at least 85% identical to SEQ ID NO: 75, or

c) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 77 or a sequence at least 85% identical to SEQ ID NO: 77, and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 75 or a sequence at least 85% identical to SEQ ID NO: 75, or

d) the polypeptide of formula [IV] comprises the Fc domain (Fc2) of SEQ ID NO: 77 or a sequence at least 85% identical to SEQ ID NO: 77, and

the polypeptide of formula [III] comprises the Fc domain of SEQ ID NO: 79 or a sequence at least 85% identical to SEQ ID NO: 79.

4. The antibody-like binding protein according to any one of claims 1 to 3, wherein a) one polypeptide of formula [IV] consists of the amino acid sequence

SEQ ID NO: 80, or

a sequence at least 85% identical to SEQ ID NO: 80 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 of SEQ ID NO: 80 are unaltered; and

one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 59, or a sequence at least 85% identical to SEQ ID NO: 59 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 ,

559, 560, 478, 490 of SEQ ID NO: 59 are unaltered; or

b) one polypeptide of formula [IV] consists of the amino acid sequence SEQ ID NO: 72, or a sequence at least 85% identical to SEQ ID NO: 72 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ I D NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 of SEQ ID NO: 72 are unaltered; and

one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 74, or a sequence at least 85% identical to SEQ ID NO: 74 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence

SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 , 559, 560, 478, 490 of SEQ ID NO: 74 are unaltered; or

c) one polypeptide of formula [IV] consists of the amino acid sequence SEQ ID NO: 76, or a sequence at least 85% identical to SEQ ID NO: 76 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO:

54, and the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 in SEQ ID NO: 76 are unaltered; and

one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 74, or a sequence at least 85% identical to SEQ ID NO: 74 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 , 559, 560, 478, 490 of SEQ ID NO: 74 are unaltered; or

d) one polypeptide of formula [IV] consists of the amino acid sequence SEQ ID NO: 76, or a sequence at least 85% identical to SEQ ID NO: 76 in which the 3 CDRs of sequences SEQ ID NO: 48, WAS' and SEQ ID NO: 49 of VD1 of sequence SEQ ID NO: 54, and the 3 CDRs of sequences SEQ I D NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 and the amino acid positions 481 , 486, 498, 500, 539, 567, 568 in SEQ ID NO: 76 are unaltered; and

one polypeptide of formula [III] consists of the amino acid sequence SEQ ID NO: 78, or a sequence at least 85% identical to SEQ ID NO: 78 in which the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO: 53, and SEQ ID NO: 51 of VD4 of sequence SEQ ID NO: 52, and the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 of VD3 of sequence SEQ ID NO: 9 and the amino acid positions 473, 492, 531 ,

559, 560, 478, 490 of SEQ ID NO: 78 are unaltered.

5. The antibody-like binding protein according to any one of claims 1 to 4, comprising i) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

80; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 59,

ii) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

72, and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 74,

iii) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO:

76; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 74, or

vi) one polypeptide of formula [IV] consisting of the amino acid sequence SEQ ID NO: 76; and

one polypeptide of formula [III] consisting of the amino acid sequence SEQ ID NO: 78.

6. An antibody-like binding protein that binds specifically to human CD3e and human CD123 comprising three polypeptide chains that form two antigen-binding sites, wherein a first polypeptide has a structure represented by the formula [I]:

VD1-L VD2-L2-CL [I] and a second polypeptide chain has a structure represented by the formula [III]:

VD3-L3-VD4-L4-CHI-Fc [III];

and a third polypeptide FC3 which is the immunoglobulin hinge region and CH2, CH3 immunoglobulin heavy chain constant domains of an immunoglobulin;

wherein

c) said polypeptide of formula [I] consists of:

(iii) the amino acid sequence SEQ ID NO: 55 which comprises

• VD1 of sequence SEQ ID NO: 54,

• Li of sequence SEQ ID NO: 56,

• VD2 of sequence SEQ ID NO: 10,

· l_2 of sequence SEQ ID NO: 56,

• CL of sequence SEQ ID NO: 18,

or

(iv) a sequence at least 85% identical to SEQ ID NO: 55 in which

• the 3 CDRs of sequences SEQ ID NO: 48, 'WAS' and SEQ ID NO:

49 of VD1 of sequence SEQ ID NO: 54, are unaltered and

• the 3 CDRs of sequences SEQ ID NO: 1 1 , 'KVS' and SEQ ID NO: 8 of VD2 of sequence SEQ ID NO: 10 are unaltered;

d) said polypeptide of formula [III] consists of:

(iii) the amino acid sequence SEQ ID NO: 67 which comprises:

· VD3 of sequence SEQ ID NO: 9,

• l_3 which consists of 0 amino acid,

• VD4 of sequence SEQ ID NO: 52,

• l_4 which consists of 0 amino acid,

• CH1 of sequence SEQ ID NO: 19, and

• Fc of sequence SEQ ID NO: 68 wherein Xi is Y, X2 is C, X3 is W, X4 is L, X5 is Y, and X6 is H and X7 is Y, or X6 is R and X7 is F, or

(iv) a sequence at least 85% identical to SEQ ID NO : 67 in which

· the 3 CDRs of sequences SEQ ID NO: 50, SEQ ID NO:53, and SEQ

ID NO: 51 of VD4 of sequence SEQ ID NO: 52, are unaltered and

• the 3 CDRs of sequences SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID

NO: 7 of VD3 of sequence SEQ ID NO: 9 are unaltered, and

• the amino acids Xi, X2, X3, X4, X5, Χβ and X7 are as defined above in b)(i),

and wherein:

- the polypeptide formula [I] and the polypeptide of formula [III] form a cross-over light chain-heavy chain pair,

- the polypeptide of formula [III] heterodimerizes with the third polypeptide through its Fc domain

- said third polypeptide Fc3 consists of SEQ ID NO: 69 or a sequence at least 85% identical to SEQ ID NO: 69, wherein the amino acid positions 129, 146, 148, 187, 215, 216 of SEQ ID NO: 69 are unaltered.

7. The antibody-like binding protein according to claim 6, comprising

- one polypeptide of formula [I] consisting of SEQ ID NO: 55,

- one polypeptide of formula [III] consisting of SEQ ID NO: 65, and

- Fc3 consisting of SEQ ID NO: 69.

8. A pharmaceutical composition comprising an antibody-like binding protein according to any one of claims 1 to 7 and a pharmaceutically acceptable carrier.

9. An antibody-like binding protein according to any one of claims 1 to 7 or a pharmaceutical composition according to claim 8 for use as a medicament.

10. An antibody-like binding protein according to any one of claims 1 to 7 or a pharmaceutical composition according to claim 8 for use for the treatment of cancer.

1 1 . An antibody-like binding protein or pharmaceutical composition for the use according to claim 9, wherein the cancer is a haematological cancer.

12. A method of treating or preventing a disease or disorder comprising administering to a subject in need thereof a therapeutically effective amount of an antibody-like binding protein according to any one of claims 1 to 7 or a pharmaceutical composition according to claim 8.

13. An isolated nucleic acid comprising a sequence encoding an antibody-like binding protein according to any one of claims 1 to 7.

14. A host cell which has been transformed by a nucleic acid according to claim 13.

15. A kit comprising

a) at least one antibody-like binding protein as defined according to any one of claims 1 to 7,

b) optionally packaging material, and

c) optionally a label or packaging insert contained within said packaging material indicting that said antibody-like binding protein is for effective for treating cancer or for use for the treatment of cancer.