CD20 and CD22 targeted antigen binding molecules for use in proliferative disorders

Abstract

The present invention provides a CD20- and CD22-targeting antigen-binding molecule, characterized in that it comprises a first and a second domain that bind to CD20 and CD22, respectively, a third domain that binds to an extracellular epitope of the human and Macaca CD3ε chain, and an optional fourth domain that is an Fc format. Furthermore, the present invention provides a polynucleotide encoding this antigen-binding molecule, a vector comprising this polynucleotide, a host cell that expresses this antigen-binding molecule, and a pharmaceutical composition comprising the same.

Description

[Technical field]

[0001] The present invention relates to biotechnology products and methods, in particular to CD20- and CD22-targeting antigen-binding molecules, their preparation, and uses thereof. [Background technology]

[0002] Bispecific molecules useful in immuno-oncology can be antigen-binding polypeptides such as antibodies, for example, IgG-like bispecific antibodies (i.e., full-length bispecific antibodies), or non-IgG-like bispecific antibodies that are not full-length antigen-binding molecules. Full-length bispecific antibodies usually retain the normal monoclonal antibody (mAb) structure of two Fab arms and one Fc region, except that the two Fab sites bind different antigens. Non-full-length bispecific antibodies may completely lack the Fc region. These include chemically linked Fabs consisting of only the Fab region, as well as various types of bivalent and trivalent single-chain variable fragments (scFv). There are also fusion proteins that mimic the variable domains of two antibodies. An example of such a format is the Bispecific T Cell Engager (BiTE®) (Yang, Fa; Wen, Weihong; Qin, Weijun (2016). "Bispecific Antibodies as a Development Platform for New Concepts and Treatment Strategies". International Journal of Molecular Sciences. 18(1):48).

[0003] Exemplary bispecific antibody-derived molecules, such as BiTE® molecules, are recombinant protein constructs composed of two flexibly linked antibody-derived binding domains. One binding domain of the BiTE® molecule is specific for a selected tumor-associated surface antigen on target cells; the second binding domain is specific for CD3, a subunit of the T cell receptor complex on T cells. Due to their special design, BiTE® antigen-binding molecules are uniquely suited to transiently link T cells to target cells and simultaneously potently activate the intrinsic cytolytic capacity of T cells against target cells. Significant further development of the first generation of BiTE® molecules deployed in the clinic as AMG 103 and AMG 110 (see WO 99 / 54440 and WO 2005 / 040220) provided bispecific antigen-binding molecules that bind to a context-independent epitope at the N-terminus of the CD3 epsilon chain (WO 2008 / 119567). BiTE® molecules binding to this selected epitope not only do not show cross-species specificity for human and Macaca, or Callithrix jacchus, Saguinus oedipus or Saimiri sciureus CD3ε chains, but also do not show the same degree of non-specific activation of T cells as observed with previous generation T cell engaging antibodies, since they recognize this specific epitope (instead of the CD3 binding epitope already described in the bispecific T cell engaging molecules). This reduction in T cell activation is associated with a reduced or reduced T cell redistribution in patients, the latter identified as a risk of side effects, for example with pasutuximab.

[0004] The antibody-based molecules described in WO 2008 / 119567 are characterized by rapid clearance from the body; therefore, although they can reach most parts of the body quickly, their in vivo application may be limited by their short persistence in vivo. On the other hand, their concentration in the body can be quickly adapted and fine-tuned. Since the small single-chain molecules have a short in vivo half-life, continuous administration by continuous intravenous infusion is used to achieve a therapeutic effect. However, bispecific antigen-binding molecules are available with more favorable pharmacokinetic properties, including a longer half-life, as described in WO 2017 / 134140. A long half-life is usually useful in in vivo applications of immunoglobulins, for example for patient compliance, and is particularly useful in in vivo applications involving small-sized antibody fragments or constructs.

[0005] One of the ongoing challenges in antibody-based immuno-oncology is tumor escape. Such tumor escape occurs when the immune system, even when induced or induced by some antibody-based immunotherapy, cannot sufficiently eradicate tumors that have accumulated genetic and epigenetic modifications and use some mechanisms to be victorious in the immune editing process (Keshavarz-Fathi,Mahsa;Rezaei,Nima(2019)"Vaccines for Cancer Immunotherapy"). In general, four mechanisms are known that impede effective anti-tumor immune responses: (1) defective tumor antigen processing or presentation, (2) lack of activation mechanisms, (3) inhibitory mechanisms and immune suppression states, and (4) resistant tumor cells. In particular, with regard to the first mechanism, tumor antigens may exist in new forms due to genetic instability, tumor mutations, and escape from the immune system. Epitope-negative tumor cells remain hidden and are therefore resistant to immune rejection. These are expressed following the elimination of epitope-positive tumor cells, similar to Darwin's theory of natural selection.As a result, antibody-based immunotherapy against antigens on tumor cells becomes ineffective when such tumor cells no longer express the respective antigens due to tumor escape.Said antigen loss is understood herein as the driving force of tumor escape, and therefore is used interchangeably.Therefore, there is a need to provide an improved antibody-based cancer immunotherapy that addresses the problem of antigen loss in order to effectively prevent tumor escape.

[0006] Moreover, despite the preclinical and clinical successes achieved so far with antibody-based immunotherapy, significant limitations remain, such as differences in response between individuals and between cancer types. Not all patients respond to therapy at available safe doses, because dose-limiting toxicity can be a factor limiting the effectiveness of antibody-based immunotherapeutics. Therefore, there is also a need to reduce the dose-limiting toxicity in antibody-based immunotherapeutics, making such therapy available to more patients suffering from various proliferative diseases.

[0007] Another challenge to the widespread use of cancer immunotherapy for T cell-engaging bispecific molecules is the availability of suitable targets (Bacac et al., Clin Cancer Res; 22(13)July 1,2016). For example, solid tumor targets may be overexpressed on tumor cells, but expressed at lower but significant levels on non-malignant primary cells in key tissues. In nature, according to Bacac et al., T cells may distinguish between cells with high and low antigen expression by T cell receptors (TCRs) that can achieve relatively low affinity, but still high avidity binding to target cells expressing sufficiently high levels of the target antigen. Therefore, T cell-engaging bispecific molecules that can facilitate the same and thus maximize the period between the killing of cells with high and low target expression are highly desirable. One approach that has been discussed in the art is the use of dual targeting of two antigens on the same cells to improve target selectivity over normal tissues that express only one or low levels of both target antigens. This effect is believed to depend on the binding activity component mediated by the simultaneous binding of bsAb to both antigens on the same cell. For such dual targeting, several multispecific monoclonal antibodies (mAbs) or other immune constructs are known in the art. WO 2014 / 116846 teaches a multispecific binding protein comprising a first binding site that specifically binds to a target cell antigen, a second binding site that specifically binds to a cell surface receptor on an immune cell, and a third binding site that specifically binds to a cell surface regulator on an immune cell. US 2017 / 0022274 discloses a trivalent T cell redirecting complex comprising a bispecific antibody, which has two binding sites for tumor-associated antigens (TAA) and one binding site for T cells.Although various multispecific antibodies or antibody fragments are known in the art, some of which address T cells, no CD20 and CD22 targeted bispecific molecule employing the mechanisms of (preferably single chain) bispecific T cell engaging molecules has been proposed so far, addressing the need to overcome antigen deficiency / tumor evasion and to reduce dose-limiting toxicity in antibody-based immunotherapy, while addressing the need to effectively redirect T cells with one stable, ready-to-use therapeutic system. [Prior art documents] [Patent documents]

[0008] [Patent Document 1] International Publication No. 99 / 54440 Brochure [Patent Document 2] International Publication No. 2005 / 040220 Brochure [Patent Document 3] International Publication No. 2008 / 119567 Brochure [Patent Document 4] International Publication No. 2017 / 134140 Brochure [Patent Document 5] International Publication No. 2014 / 116846 Brochure [Patent Document 6] US Patent Application Publication No. 2017 / 0022274 [Non-patent literature]

[0009] [Non-Patent Document 1] Yang, Fa; Wen, Weihong; Qin, Weijun (2016). “Bispecific Antibodies as a Development Platform for New Concepts and Treatment Strategies”. International Journal of Molecular Sciences. 18(1):48 [Non-Patent Document 2] Keshavarz-Fathi, Mahsa; Rezaei, Nima (2019) “Vaccines for Cancer Immunotherapy” [Non-Patent Document 3] Bacac et al., Clin Cancer Res;22(13)July 1,2016 Summary of the Invention [Means for solving the problem]

[0010] In view of the above-described needs, it is an object of the present invention to provide a CD20 and CD22 targeted antigen binding molecule (typically a polypeptide, e.g. a T cell engaging bispecific molecule) for use in the treatment of a particular condition, which is particularly suitable for binding two antigens on target cells associated with said particular condition and simultaneously binding one antigen on effector cells. The molecule should further exhibit high productivity, stability and activity. Thus, the present invention provides a CD20 and CD22 targeted bispecific antigen binding molecule, characterized in that it comprises a first domain that binds to CD20 as a first target cell surface antigen (TAA), a second domain that binds to CD22 (a second TAA), a third domain that binds to an extracellular epitope of the CD3ε chain of human and non-human (e.g. Macaca), and preferably a fourth domain that is a specific Fc format that regulates the half-life of the molecule. Preferably, these domains are binding domains composed of a VH domain and a VL domain, respectively, in the amino to carboxyl direction, which are flexible but short peptides. The tidolinker is a binding domain that links the VL of the first binding domain to the VH of the second binding domain. Surprisingly, the activity of the molecules of the present invention against target cells associated with a particular disease can be preserved without steric hindrance between the first and second binding domains and without the need to provide a long linker that would be disadvantageously susceptible to degradation, cleavage, or the like compared to the shorter linkers provided immediately. At the same time, the molecules are well producible and show good product identity. Furthermore, the present invention provides a polynucleotide encoding an antigen-binding molecule, a vector comprising the polynucleotide, and a host cell expressing the construct, as well as a pharmaceutical composition comprising the same.

[0011] In a first aspect, the present invention relates to A CD20 and CD22 targeted antigen binding molecule comprising at least three binding domains, (i.) a first binding domain comprises a paratope that immunospecifically binds to CD20, the first binding domain comprising: a) CDR H1 to 3 of SEQ ID NOs: 58 to 60, and CDR L1 to 3 of SEQ ID NOs: 61 to 63; b) CDR H1 to 3 of SEQ ID NOs: 71 to 73, and CDR L1 to 3 of SEQ ID NOs: 74 to 76; c) CDR H1 to 3 of SEQ ID NOs: 84 to 86, and CDR L1 to 3 of SEQ ID NOs: 87 to 89, and d) CDR H1 to 3 of SEQ ID NOs: 97 to 99, and CDR L1 to 3 of SEQ ID NOs: 100 to 102 and a VH region comprising CDR-H1, CDR-H2, and CDR-H3 selected from the group consisting of: (ii.) the second binding domain comprises a paratope that immunospecifically binds to CD22, the second binding domain comprising: a) CDR H1 to 3 of SEQ ID NOs: 138 to 140, and CDR L1 to 3 of SEQ ID NOs: 141 to 143; b) CDR H1 to 3 of SEQ ID NOs: 151 to 153, and CDR L1 to 3 of SEQ ID NOs: 154 to 156; c) CDR H1 to 3 of SEQ ID NOs: 164 to 166, and CDR L1 to 3 of SEQ ID NOs: 167 to 169; d) CDR H1 to 3 of SEQ ID NOs: 177 to 179, and CDR L1 to 3 of SEQ ID NOs: 180 to 182; e) CDR H1 to 3 of SEQ ID NOs: 190 to 192, and CDR L1 to 3 of SEQ ID NOs: 193 to 195; f) CDR H1 to 3 of SEQ ID NOs: 203 to 205, and CDR L1 to 3 of SEQ ID NOs: 206 to 208; g) CDR H1 to 3 of SEQ ID NOs: 125 to 127, and CDR L1 to 3 of SEQ ID NOs: 128 to 130; h) CDR H1 to 3 of SEQ ID NOs: 216 to 218, and CDR L1 to 3 of SEQ ID NOs: 219 to 221, and i) CDR H1 to 3 of SEQ ID NOs: 379 to 381, and CDR L1 to 3 of SEQ ID NOs: 382 to 384 and a VH region comprising CDR-H1, CDR-H2, and CDR-H3 selected from the group consisting of: (iii.) the third binding domain comprises a paratope that immunospecifically binds to an extracellular epitope of the human and / or Macaca CD3 epsilon chain; the first, second, and third binding domains are arranged in the order of amino to carboxyl, and the first binding domain and the second binding domain are linked by a peptide linker having a length of 5 to 24, preferably 18, amino acids; It is envisaged to provide CD20 and CD22 targeted antigen binding molecules.

[0012] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, which comprises a fourth domain, the fourth domain comprising two polypeptide monomers, each of which comprises a hinge, a CH2 domain and a CH3 domain, and the two polypeptide monomers are fused to each other via a peptide linker.

[0013] In said aspect, in relation to the present invention there is provided a multispecific antigen-binding molecule, wherein said fourth domain comprises, in order from amino to carboxyl: Hinge-CH2-CH3-Linker-Hinge-CH2-CH3 It is also envisaged to provide a multispecific antigen-binding molecule comprising:

[0014] In the above aspect, it is also envisaged in relation to the present invention to provide a multispecific antigen-binding molecule, wherein each of the polypeptide monomers in the fourth domain has an amino acid sequence that is at least 90% identical to a sequence selected from the group consisting of SEQ ID NOs: 17 to 24, and preferably each of the polypeptide monomers has an amino acid sequence selected from SEQ ID NOs: 17 to 24.

[0015] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, wherein the CH2 domain comprises an intradomain cysteine ​​disulfide bridge.

[0016] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, wherein the first, second, third and optionally fourth binding domains are arranged in amino to carboxyl order.

[0017] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, wherein the antigen-binding molecule is a single chain antigen-binding molecule, preferably a multispecific scFv antigen-binding molecule.

[0018] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, wherein the first, second and third binding domains each comprise, in amino to carboxyl order, a VH domain and a VL domain.

[0019] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, wherein the peptide linker between the VH of the first binding domain and the VH of the second binding domain is selected from those having a length of 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23 or 24 amino acids, preferably 5, 6, 7, 8, 9, 10, 11 or 12 amino acids, more preferably 6 amino acids.

[0020] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, wherein the peptide linker between the VL of the first binding domain and the VH of the second binding domain is a flexible linker and comprises serine and glycine as amino acid building blocks, preferably a flexible linker comprising only serine (Ser, S) and glycine (Gly, G).

[0021] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, wherein the peptide linker between the first and second binding domain is preferably rich in small and / or hydrophilic amino acids, preferably selected from the group consisting of S(G4S)n, (G4S)n, (G4)n and (G5)n (wherein n is equal to 1, 2, 3 or 4, more preferably n is equal to 1 or 2), more preferably selected from SG4S.

[0022] In said aspect, in the context of the present invention there is provided a CD20 and CD22 targeting antigen binding molecule, comprising: The first binding domain and the second binding domain each have a) CDRs H1 to 3 of SEQ ID NOs: 58 to 60 and CDRs L1 to 3 of SEQ ID NOs: 61 to 63 in a first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 138 to 140 and CDRs L1 to 3 of SEQ ID NOs: 141 to 143 in a second binding domain; b) CDRs H1 to 3 of SEQ ID NOs: 58 to 60 and CDRs L1 to 3 of SEQ ID NOs: 61 to 63 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 151 to 153 and CDRs L1 to 3 of SEQ ID NOs: 154 to 156 in the second binding domain; c) CDRs H1 to 3 of SEQ ID NOs: 58 to 60 and CDRs L1 to 3 of SEQ ID NOs: 61 to 63 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 164 to 166 and CDRs L1 to 3 of SEQ ID NOs: 167 to 169 in the second binding domain; d) CDRs H1 to 3 of SEQ ID NOs: 58 to 60 and CDRs L1 to 3 of SEQ ID NOs: 61 to 63 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 177 to 179 and CDRs L1 to 3 of SEQ ID NOs: 180 to 182 in the second binding domain; e) CDRs H1 to 3 of SEQ ID NOs: 58 to 60 and CDRs L1 to 3 of SEQ ID NOs: 61 to 63 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 190 to 192 and CDRs L1 to 3 of SEQ ID NOs: 193 to 195 in the second binding domain; f) CDRs H1 to 3 of SEQ ID NOs: 58 to 60 and CDRs L1 to 3 of SEQ ID NOs: 61 to 63 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 203 to 205 and CDRs L1 to 3 of SEQ ID NOs: 206 to 208 in the second binding domain; g) CDRs H1 to 3 of SEQ ID NOs: 58 to 60 and CDRs L1 to 3 of SEQ ID NOs: 61 to 63 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 125 to 127 and CDRs L1 to 3 of SEQ ID NOs: 128 to 130 in the second binding domain; h) CDRs H1 to 3 of SEQ ID NOs: 58 to 60 and CDRs L1 to 3 of SEQ ID NOs: 61 to 63 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 216 to 218 and CDRs L1 to 3 of SEQ ID NOs: 219 to 221 in the second binding domain; i) CDRs H1 to 3 of SEQ ID NOs: 71 to 73 and CDRs L1 to 3 of SEQ ID NOs: 74 to 76 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 379 to 381 and CDRs L1 to 3 of SEQ ID NOs: 382 to 384 in the second binding domain; j) CDRs H1 to 3 of SEQ ID NOs: 71 to 73 and CDRs L1 to 3 of SEQ ID NOs: 74 to 76 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 203 to 205 and CDRs L1 to 3 of SEQ ID NOs: 206 to 208 in the second binding domain; k) CDRs H1 to 3 of SEQ ID NOs: 84 to 86 and CDRs L1 to 3 of SEQ ID NOs: 87 to 89 of the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 164 to 166 and CDRs L1 to 3 of SEQ ID NOs: 167 to 169 of the second binding domain; l) CDRs H1 to 3 of SEQ ID NOs: 97 to 99 and CDRs L1 to 3 of SEQ ID NOs: 100 to 102 in the first binding domain, and CDRs H1 to 3 of SEQ ID NOs: 177 to 179 and CDRs L1 to 3 of SEQ ID NOs: 180 to 182 in the second binding domain; m) CDR H1 to 3 of SEQ ID NOs: 97 to 99 and CDR L1 to 3 of SEQ ID NOs: 100 to 102 of the first binding domain, and CDR H1 to 3 of SEQ ID NOs: 190 to 192 and CDR L1 to 3 of SEQ ID NOs: 193 to 195 of the second binding domain It is also envisaged to provide a CD20 and CD22 targeting antigen binding molecule comprising a VH region comprising CDR-H1, CDR-H2, and CDR-H3, and a VL region comprising CDR-L1, CDR-L2, and CDR-L3 selected from:

[0023] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, wherein a first binding domain is capable of binding to a first target cell surface antigen CD20 and a second binding domain is simultaneously capable of binding to a second target cell surface antigen CD22, preferably, the first target cell surface antigen and the second target cell surface antigen are present on the same target cell.

[0024] In said aspect, in the context of the present invention, there is provided a CD20 and CD22 targeting antigen binding molecule according to claim 1, wherein the third binding domain comprises a) CDR H1 to 3 of SEQ ID NOs: 392 to 394, and CDR L1 to 3 of SEQ ID NOs: 395 to 397; and b) CDR H1 to 3 of SEQ ID NOs: 401 to 403, and CDR L1 to 3 of SEQ ID NOs: 404 to 406 It is also envisaged to provide a CD20 and CD22 targeting antigen binding molecule comprising a VH region comprising CDR-H1, CDR-H2, and CDR-H3, and a VL region comprising CDR-L1, CDR-L2, and CDR-L3 selected from:

[0025] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, in which the first, second and third domains, each fused by a peptide linker, are fused to a fourth domain via a peptide linker.

[0026] In said aspect, the present invention relates to a multispecific antigen-binding molecule, the antigen-binding molecule comprising, in order from amino to carboxyl, (a) the first domain; (b) a peptide linker having an amino acid sequence preferably selected from the group consisting of SEQ ID NOs: 1 to 4 and 9 to 12, preferably selected from SEQ ID NO: 11; (c) a second domain; (d) a peptide linker preferably having an amino acid sequence selected from the group consisting of SEQ ID NOs: 1 to 3; and (e) Third Domain It is also envisaged to provide a multispecific antigen-binding molecule comprising:

[0027] In said aspect, the present invention relates to a multispecific antigen-binding molecule, the antigen-binding molecule comprising, in order from amino to carboxyl, (f) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 1, 2, 3, 9, 10, 11, and 12; (g) a first polypeptide monomer of the fourth domain; (h) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 5, 6, 7, and 8; and (i) a second polypeptide monomer of the fourth domain It is also envisaged to provide a multispecific antigen-binding molecule further comprising:

[0028] In the above aspect, it is also envisaged in the context of the present invention to provide an antigen-binding molecule, wherein the first binding domain comprises a VH region and a VL region selected from SEQ ID NO: 64 as VH and SEQ ID NO: 65 as VL, SEQ ID NO: 77 as VH and SEQ ID NO: 78 as VL, SEQ ID NO: 90 as VH and SEQ ID NO: 91 as VL, SEQ ID NO: 103 as VH and 104 as VL, respectively, and the second binding domain comprises a VH region and a VL region selected from SEQ ID NO: 144 as VH and SEQ ID NO: 145 as VL, SEQ ID NO: 157 and 158, SEQ ID NO: 172 and 173, SEQ ID NO: 183 and 184, SEQ ID NO: 196 and 197, SEQ ID NO: 209 and 210, SEQ ID NO: 131 and 132, and SEQ ID NO: 385 and 386, respectively.

[0029] In said aspect, it is also envisaged in the context of the present invention to provide an antigen-binding molecule, wherein the first binding domain comprises an scFv sequence selected from the group consisting of SEQ ID NOs: 66, 79, 92, and 105, and the second binding domain comprises an scFv sequence selected from the group consisting of SEQ ID NOs: 146, 159, 172, 185, 198, 211, 133, 224, and 387.

[0030] In said aspect, it is also envisaged in the context of the present invention to provide a multispecific antigen-binding molecule, which comprises a first (CD20) and a second (CD22) target binding domain together with a third effector (CD3) binding domain and a fourth domain that confers half-life extension, wherein the three binding domains linked together and the fourth domain have a sequence selected from the group consisting of SEQ ID NOs: 238, 248, 258, 268, 278, 288, 308, 318, 328, 338, 348, 368 and 378.

[0031] In a second aspect, it is further envisaged in the context of the present invention to provide a polynucleotide encoding the antigen-binding molecule of the present invention.

[0032] In a third aspect, it is also envisaged in relation to the present invention to provide a vector comprising a polynucleotide of the invention.

[0033] In a fourth aspect, it is further envisaged in connection with the present invention to provide a host cell transformed or transfected with a polynucleotide or vector of the invention.

[0034] In a fifth aspect, it is also envisaged in the context of the present invention to provide a process for producing an antigen-binding molecule of the present invention, the process comprising culturing a host cell of the present invention under conditions allowing expression of the antigen-binding molecule, and recovering the produced antigen-binding molecule from the culture.

[0035] In a sixth aspect, it is further contemplated in connection with the present invention to provide a pharmaceutical composition comprising an antigen-binding molecule of the present invention or an antigen-binding molecule produced according to the process of the present invention.

[0036] In said embodiment, it is also envisaged in the context of the present invention that the pharmaceutical composition is stable at about -20°C for at least 4 weeks.

[0037] In the context of the present invention, it is further envisaged to provide an antigen-binding molecule of the present invention, or an antigen-binding molecule produced according to the process of the present invention, for use in the prevention, treatment or amelioration of a disease selected from a proliferative disease, a neoplastic disease, a cancer or an immune disorder.

[0038] In said embodiment, it is also envisaged in the context of the present invention that CD20xCD22 targeting antigen binding molecules are used in the treatment of non-Hodgkin's lymphoma. [Brief description of the drawings]

[0039] [Figure 1-1] 48-hour FACS-based cytotoxicity assay of CD20 and CD22 dual-targeting antigen binding molecules with human CD20 and CD22 double-positive human cell line Oci-Ly 1 (A), human CD20 single-positive human cell line Oci-Ly 1 (CD22 knockout clone #A1) (B), and CD22 single-positive human cell line Oci-Ly 1 (CD20 knockout clone #A5) (C) as target cells, and panT (E:T ratio 10:1) as effector cells. EC50 values ​​are determined by a four-parametric logistic regression model for evaluation of sigmoidal dose-response curves with a fixed Hill slope. [Figure 1-2]48-hour FACS-based cytotoxicity assay of CD20 and CD22 dual-targeting antigen binding molecules with human CD20 and CD22 double-positive human cell line Oci-Ly 1 (A), human CD20 single-positive human cell line Oci-Ly 1 (CD22 knockout clone #A1) (B), and CD22 single-positive human cell line Oci-Ly 1 (CD20 knockout clone #A5) (C) as target cells, and panT (E:T ratio 10:1) as effector cells. EC50 values ​​are determined by a four-parametric logistic regression model for evaluation of sigmoidal dose-response curves with a fixed Hill slope. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0040] In the context of the present invention, there is provided a CD20- and CD22-targeted antigen-binding molecule comprising at least three binding domains, wherein a first and a second binding domain, preferably in an amino to carboxyl orientation, can simultaneously target CD20 and CD22, and a third binding domain binds to an extracellular epitope of the human and / or macaque CD3ε chain on effector cells, which are T cells.

[0041] In the context of the present invention, it is a surprising discovery that the T cell engaging CD20 and CD22 targeted antigen binding molecules of the present invention, having a selected combination of CD20 and CD22 targeting binders, show excellent yield, stability, and balanced activity between the two targeting binders. This improves both practical aspects such as production and storage capacity, and preferably reliable efficacy. In this regard, the molecules of the present invention are found to show HIC elution gradients, as demonstrated herein, that are typically greater than 15, preferably greater than 20, or even greater than 25. However, molecules according to the general structure of the present invention, not including the specific binder selection described herein, typically show lower values, indicating a lower homogeneity of the product. Even more remarkable is the yield, as an indicator of overall productivity, which is typically greater than 10 mg / L, preferably greater than 15 mg / L, or even greater than 20 mg / L of monomer (i.e., desired product). In contrast, other molecules of the general format on which the molecules described herein are based typically do not reach a yield of greater than 10 mg / L. As a further indication of product quality, the symmetry of the monomer peak in size exclusion chromatography (SEC) is typically improved for molecules comprising the specific binder selection of the present invention. Such peak symmetry is preferably below a value of 1.4, more preferably below a value of 1.35 or less. As will be appreciated by those skilled in the art, values ​​around 1 are typically preferred. However, other molecules following the general format on which the molecules of the present invention are based typically do not reach values ​​below 1.4. Furthermore, the molecules of the present invention typically show good activity against cells expressing both targets CD20 and CD22. Thus, for molecules comprising the specific selection of anti-CD20 and anti-CD22 binders claimed herein, the EC50 values ​​observed are typically surprisingly low. Thus, the molecules of the present invention typically show EC50 values ​​against CD20-CD22 double positive target cells (e.g., Oci-Ly 1 cells) below 20 pM, preferably below 15 pM, even more preferably below 10 pM.Other molecules that follow the general format on which the molecules of the invention are based typically exhibit EC50 values ​​above 20 pM under corresponding conditions. Thus, higher efficacy can be attributed to the molecules of the invention.

[0042] In addition, the molecules of the present invention are preferably suitable for targeting two (different) antigens on one target cell, such as a cancer cell, and in contrast, do not target non-cancer cells well, fulfilling the surprising feature of the underlying general type of molecule. By being able to simultaneously address two target antigens, (a) the possibility of targeting a target cell, such as a cancer cell, is greatly increased when such a target cell undergoes antigen deficiency, and thus tumors tend to avoid effective antitumor therapy, since there remains one effective antigen to target on cells that have undergone antigen avoidance, and (b) the possibility of targeting a target cell associated with a disease instead of a physiological cell is greatly increased when two TAAs that are typically associated with a target cell associated with a disease instead of a physiological cell are selected. In this regard, CD20 and CD22 targeting antigen binding molecules are envisaged herein, which not only prevent antigen avoidance, for example, in the tumor environment, but also further widen the therapeutic window by addressing cells that have a pattern of two antigens that are typically associated with a particular disease. Thus, biological tissues in which cells express only one of the two targets are not addressed by immediate dual targeting antigen binding molecules. In particular, the selective gap can be achieved by a dual targeting molecule, for example in the format as described herein, having a bispecific entity comprising a target binding domain (or binder, used interchangeably throughout this disclosure) and a CD3 binder, a further target binder, and an optional half-life extending domain (e.g., an scFc domain). The dual targeting antigen binding molecules described herein are typically characterized by an EC50 value of less than 100 pM against cells positive for both targets, preferably less than 50 pM, more preferably less than 30 pM, and even more preferably about 10 pM or less, whereas such dual targeting molecules typically exhibit significantly higher EC50 values ​​(e.g., at least 50 pM, 100 pM, 250 pM, or even 500 pM or more) when used with single targeting cells.This finding suggests that the CD20 and CD22 targeting molecules of the present invention advantageously can be used to specifically address pathogenic target cells expressing both targets, and do have a selectivity gap in terms of activity of at least 10-fold, preferably at least 20-fold, or even 30-fold, that can induce T cell-mediated cytotoxicity by simultaneous binding of the molecules. Off-target toxicity and associated side effects may thereby be reduced, providing a safer treatment based on the concepts described herein. Thus, the T cell-engaging CD20 and CD22 targeting antigen binding molecules of the present invention, which are typically single chains, provide improved efficacy and safety with respect to existing bispecific antibodies or antigen binding molecules that bind to T cells. The advantageous properties are preferably achieved by the fact that the first and second binding domains of the CD20 and CD22 targeting antigen binding molecules can maintain their biological activity independently of each other (i.e., can bind to their respective targets without being sterically hindered by the targets to which the respective other binding domains and / or respective other target binding bodies are bound). Preservation of biological activity is preferably achieved by (a) the amino to carboxyl VH-VL arrangement of both binding domains and / or (b) careful selection of the linker linking the first and second binding domains. The linker must have a length that ensures both biological activity of both binding domains and sufficient (chemical) stability of the construct. Surprisingly, relatively short peptide linkers of about 5-24, preferably 5-18, more preferably 6 or 12 amino acids in length meet both requirements. Preferably, such linkers are rich in small or hydrophilic amino acids such as Gly and Ser, since such a composition preferably provides flexibility. Such flexibility, in turn, preferably allows interaction of the respective binding domains independently of the other binding domains of the CD20- and CD22-targeted antigen binding molecules according to the invention.At the same time, it is surprising that even such short and preferably flexible peptide linkers typically provide sufficient spatial separation between the first and second binding domains such that both domains retain their biological activity required to have a therapeutically useful molecule in the context of the present invention. A further advantage of such short linkers disclosed in the context of the present invention is that interchain mispairing is preferably prevented compared to longer linkers.

[0043] The above specific findings underlying the present invention are surprising in view of the teachings of the prior art. For example, Liu et al. showed that the longer the linker between the peptides, the better the independent folding and biological activity of the two molecules are maintained (Liu ZG, Lin JB, Du W, et al. Anti-proteolysis study of recombinant In-UK fusion protein in CHO cell. Prog Biochem Biophys 2005;32:544-50). An excessively short linker between the binding domains (preferably scFv binding domains) adversely affects protein folding due to space occupation, and an excessively long linker enhances the antigenicity of the scFv antibody and also affects the functionality and activity of the scFv antibody. Xu et al. teach that sufficient length and certain sequence properties are important factors for the two half molecules to have enough free space to perform their functions, and avoiding the formation of a-helices and b-sheets is important for stability (Xue F, Gu Z, Feng JA. LINKER: a web server to generate peptide sequences with extended conformation. Nucleic Acids Res 2004;32:W562-5). Therefore, those skilled in the art who aim to maintain the distance between binding domains will typically intend to employ rigid linkers characterized by helical structures or rich in prolines. However, the length of the rigid linker also has a large impact on the biological activity of the protein. McCormick et al. investigated the application of rigid peptide linkers (Ala-Pro)n (10-34 aa) in interferon-gamma-gp120 fusion proteins (McCormick A, Thomas M, Heath A. Immunization with an interferon-gamma-gp120 fusion protein induces enhanced immune responses to human immunodeficiency virus gp120. J Infect Dis. 2001;184:1423-1430).With a short linker of 10 aa, the fusion protein had relatively low biological activity of interferon-γ. By lengthening the linker, the biological activity of the fusion protein was gradually improved, with the longest linker of 34 residues showing a peak activity of 88% of free interferon-γ. Furthermore, in some cases, even when a flexible or rigid linker was inserted, the decrease in biological activity still could not be overcome due to steric hindrance between the domains (Bai Y, Ann DK, Shen WC. Recombinant granulocyte colony-stimulating factor-transferrin fusion protein as an oral myelopoietic agent. Proc Natl Acad Sci US A.2005;102:7292-7296).

[0044] Considering the obstacles known in the art, the skilled person is encouraged to avoid short flexible or even rigid linkers and turns to longer rigid linkers, where "long" can be understood from the art as about 30 amino acids, preferably including proline. Based on this information, the skilled person will preferably use state-of-the-art modeling techniques to model the first and second binding domains connected by a peptide linker to ascertain which linker length should be taken and which linker length should be avoided. If the linker is a flexible linker rich in Ger and Ser, a linker length of 30 amino acids typically provides a fairly large space (typically at least 70 Å, more typically at least 80 Å) between the first and second binding domains, which the skilled person will consider as a safe size to accommodate the second target cell surface antigen (TAA2 CD22) to promote binding by the second binding domain of the CD20 and CD22 targeting antigen binding molecule. In the context of the present invention, it is important to note that the first binding domain (i.e. the N-terminal binding domain) is relatively easily accessible since there is only one adjacent binding domain, which may cause steric hindrance upon binding to the target, whereas the second binding domain is connected to the first binding domain in the N-direction.

[0045] Typically, when a SGGGGS linker is modeled between two target binding domains that are scFvs (a (GGGGS) linker is modeled between the VH and VL in the binding domains, respectively), when the first binding domain (e.g. anti-MSLN binding domain) is fixed, and when three possible predicted conformations (with the linker swinging at different orthogonal) are applied (linker conformations 1, 2 and 3, respectively), a complete clash is observed for linker position 3, and no clash is observed for positions 1 and 2. However, this space is typically much larger when the CDRs are targeted to the CD20 and CD22 targets according to the present invention. Based on where it is preferably located within the second binding domain of the fused antigen binding molecule, it is still not enough to accommodate TAA2. Therefore, this result strongly indicates the need for a longer linker between the two target binding domains. If one skilled in the art used the size of the target EpCAM as a guide, one would expect a better linker to have preferably at least about 30 residues, less preferably at least 20 residues (i.e., a preferred distance of 70A divided by 3.8 per aa). Thus, the lack of space led to the short linker solution such as the SGGGS linker. solution) and its short multiplicity (e.g., S(G4S)2 and S(G4S)2 between the two target binding domains according to the present invention) is unfavorable and therefore a non-obvious choice for this structure of target conjugate in CD20 and CD22 targeted antigen binding molecules (particularly dual targeting BiTE® molecules). The same is true for the 12 aa linker, which typically confers a small maximum available space of about 35 Å, which may be up to about 50 Å depending on the circumstances, so that typical targets to be bound, which are at least about 45, 50, 55, 60, 65, 70, 75, 80 or 85 Å in size, would not be safely accommodated.Similarly, an 18 aa long linker (e.g., SGGGGSGGGGSGGGGSGG) with a maximum available space between the binding domains in the structures disclosed herein of 60 Å or less, typically 55 Å or less, e.g., 54-60 Å, would likely preclude binding to a second TAA2 of exemplary size 45-70 Å. In contrast, a 30 aa long linker would typically provide a maximum space of 84-94 Å, thus allowing the target binder to safely bind to an exemplary target of about 45-70 Å. Therefore, a person skilled in the art would have selected a linker of at least more than 18 aa in length to ensure binding of the second TAA2, such as in HLE dual BiTE®, an exemplary CD20- and CD22-targeted antigen binding molecule according to the present invention. It should be noted that the above discussion is based on a flexible linker with a high content of Ser and / or Gly. The skilled person will appreciate that a less flexible linker may require more amino acids to ensure sufficient length to maintain the distance between two adjacent target binding domains according to the present invention in order to maintain the biological function of said target binding domains.

[0046] It is particularly envisaged in the context of the present invention that CD20 and CD22 targeted antigen binding molecules addressing two different target cell surface antigens are thereby highly specific for their target cells and therefore preferably safe in their therapeutic use, as demonstrated in cynomolgus monkey toxicity studies.

[0047] The B lymphocyte antigen CD20 or CD20 is expressed on the surface of all B cells beginning at the pre-B stage (CD45R+, CD117+) and gradually increasing in concentration until maturity. CD22 or cluster of differentiation 22 is a molecule that belongs to the SIGLEC family of lectins. This molecule is found on the surface of mature B cells and to a lesser extent on some immature B cells.

[0048] Furthermore, it is envisaged that optionally but advantageously in the context of the present invention, the CD20 and CD22 targeting antigen binding molecules are provided with a fourth domain (typically an scFc domain, i.e., HLE), allowing the antigen binding molecules to be administered intravenously no more than once every week, once every two weeks, once every three weeks or even once every four weeks, or even less frequently.

[0049] To determine the epitopes of preferred CD20 and CD22 targeting antigen binding molecules according to the present invention, for example directed to the CD20 epitope, mapping was performed as described herein. The extracellular region of human CD20 protein was divided into five parts: (1) extracellular loop 1, designated E1 (ECL1, amino acids 72-84, see references in Example 17), and extracellular loop 2, designated E2 (ECL2). Extracellular loop 1 (E1) was further divided into two subparts, designated E1A (aa72-79) and E1B (aa80-84). Extracellular loop 2 (E2, aa142-188) was further divided into four subparts, designated E2A (aa142-161), E2B (aa162-166), E2C (aa167-175), and E2D (aa176-188). Surprisingly, it has been discovered that CD20 antigen binding molecules exhibit preferably higher cytotoxic activity when bound to (i) the E1A and E2B and E2C epitopes, or (ii) the E2A and E2B epitopes, in both single and dual targeting. Accordingly, for epitope characterization, the extracellular region of human CD22 protein was divided into seven parts: V (aa 20-142 as defined by Uniprot P20273+RPFP), C2-1 (aa 143-241 as defined by Uniprot P20273+LNVKHT), C2-2 (aa 242-330 as defined by Uniprot P20273+VQYA), C2-3 (aa 331-418 as defined by Uniprot P20273+YP), C2-4 (aa 419-504 as defined by Uniprot P20273+VQYA), C2-5 (aa 505-592 as defined by Uniprot P20273+KAWTLEVLYA), and C2-6 (aa 600-621 as defined by Uniprot P20273+KAWTLEVLYA). Surprisingly, it was found that CD22 antigen-binding molecules exhibit preferably higher cytotoxic activity when bound to the C2-1 epitope, both in single targeting and dual targeting.

[0050] It is particularly surprising that the multispecific antigen-binding molecule according to the present invention can preferably bind simultaneously to two different targets, despite the short linker between the target-binding domains. Simultaneous binding has been demonstrated for several targets herein. However, this is surprising considering the typical distance between targets. For example, CD20 contains two small extracellular domains of only 13 aa (E1) and 47 aa (E2). In contrast, CD22 contains an extracellular domain of 7 Ig domains long with 676 aa. However, despite the significant difference in extracellular size and composition, the multispecific antigen-binding molecule according to the present invention can successfully address both TAAs CD20 and CD22 at the same time, with the advantage of high efficacy and low toxicity. This is preferably achieved.

[0051] It is envisaged in the context of the present invention that preferred multispecific antigen-binding molecules not only exhibit a favourable ratio of cytotoxicity and affinity, but also exhibit sufficient stability properties to facilitate practical handling when formulating, storing and administering said constructs. Sufficient stability is characterized by a high monomer content (i.e. non-aggregated and / or non-associated native molecules) after standard preparation, e.g. at least 65%, more preferably at least 70%, even more preferably at least 75% as determined by preparative size-exclusion chromatography (SEC). Also, the turbidity measured as light absorption at 340 nm, e.g. at a concentration of 2.5 mg / ml, should preferably be 0.025 or less, more preferably 0.020 or less, to conclude an essential lack of undesired aggregates. Advantageously, the high monomer content is maintained after incubation in stress conditions, such as freeze / thaw or incubation at 37°C or 40°C. Furthermore, multispecific antigen-binding molecules according to the invention typically have thermal stability that is at least comparable to or even greater than that of bispecific antigen-binding molecules that have only one target-binding domain but otherwise comprise a CD3-binding domain and an optional half-life-extending scFc domain (i.e., are structurally less complex). One skilled in the art would expect that more structurally complex protein-based molecules would be less susceptible to thermal and other degradation (i.e., less thermally stable).

[0052] Thus, the present invention provides a CD20 and CD22 targeting antigen binding molecule, comprising: (i.) a first binding domain that specifically binds to a first target cell surface antigen (a selected anti-CD20 binder); (ii.) a second binding domain that specifically binds to a second target cell surface antigen (a selected anti-CD22 binder); (iii.) a third binding domain that binds to an extracellular epitope of the human and / or macaque CD3 epsilon chain, the first, second, and third binding domains being arranged in amino to carboxyl order, and the first and second binding domains being linked by a peptide linker having a length of 5 to 25 (preferably 5 to 18 or 6 to 16) amino acids; and optionally (iv.) a fourth domain comprising two polypeptide monomers each comprising a hinge, a CH2, and a CH3 domain, the two polypeptide monomers being fused to each other via a peptide linker. The present invention provides a CD20 and CD22 targeting binding domain molecule comprising:

[0053] A general requirement for the CD20- and CD22-targeted bispecific antigen binding molecules of the invention is that one target binding domain must be positioned adjacent to the N-terminus of an effector CD3-binding domain in order to act as a bispecific entity and thereby form a cytolytic synapse between (preferably double positive) target cells and effector T cells.

[0054] The term "polypeptide" is understood herein as an organic polymer comprising at least one continuous unbranched amino acid chain. In the context of the present invention, polypeptides comprising multiple amino acid chains are also envisaged. The amino acid chain of a polypeptide typically comprises at least 50 amino acids, preferably at least 100, 200, 300, 400 or 500 amino acids. In the context of the present invention, it is also envisaged that the amino acid chain of the polymer is linked to an entity that is not composed of amino acids.

[0055] The term "antigen-binding polypeptide" according to the present invention is preferably a polypeptide that immunospecifically binds to its target or antigen. It typically comprises a domain comprising or derived from the heavy chain variable region (VH) and / or the light chain variable region (VL) of an antibody. A polypeptide according to the present invention comprises the minimum structural requirements of an antibody that allows immunospecific target binding. This minimum requirement may be defined, for example, by the presence of at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 of the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 of the VH region), preferably by the presence of all six CDRs. Thus, a T cell engaging polypeptide may be characterized by the presence of three or six CDRs in either or both binding domains, where (and in what order) these CDRs are located within the binding domains is known to the skilled person. Typically, an "antigen-binding molecule" is understood as an "antigen-binding polypeptide" in the context of the present invention.

[0056] Alternatively, in the context of the present invention, an antigen-binding polypeptide corresponds to an "antibody construct", which typically refers to a molecule whose structure and / or function is based on that of an antibody (e.g. a full-length or whole immunoglobulin molecule). Thus, an antigen-binding molecule may bind to its specific target or antigen and / or is derived from the variable heavy (VH) and / or variable light (VL) domains of an antibody or a fragment thereof. Furthermore, in the present specification, a domain that binds to a binding partner according to the present invention is understood as a binding domain of an antigen-binding molecule according to the present invention. Typically, a binding domain according to the present invention comprises the minimum structural requirements of an antibody that allow target binding. This minimum requirement may for example be defined by the presence of at least three light chain CDRs (i.e. CDR1, CDR2 and CDR3 of the VL region) and / or three heavy chain CDRs (i.e. CDR1, CDR2 and CDR3 of the VH region), preferably all six CDRs. An alternative approach to define the minimal structural requirements of an antibody is the definition of the epitope of the antibody within the structure of a specific target, each of which is a protein domain of the target protein that constitutes an epitope region (epitope cluster), or by reference to specific antibodies that compete with the epitope of the defined antibody. Antibodies on which the constructs of the present invention are based include, for example, monoclonal antibodies, recombinant antibodies, chimeric antibodies, deimmunized antibodies, humanized antibodies, and human antibodies.

[0057] The binding domain of the antigen-binding molecule of the present invention may, for example, comprise the CDRs of the above-referenced groups. Preferably, these CDRs are contained within the framework of an antibody light chain variable region (VL) and an antibody heavy chain variable region (VH); however, it is not necessary to contain both. Fd fragments, for example, have two VH regions and often retain some antigen-binding function of the intact antigen-binding domain. Further examples of formats of antibody fragments, antibody variants, or binding domains include (1) a Fab fragment, which is a monovalent fragment having the VL, VH, CL, and CH1 domains; (2) an F(ab')2 fragment, which is a bivalent fragment having two Fab fragments linked by a disulfide bridge at the hinge region; (3) an Fd fragment having two VH and CH1 domains; (4) an Fv fragment having the VL and VH domains of a single arm of an antibody, (5) a dAb fragment having a VH domain (Ward et al., (1989) Nature 341:544-546); (6) isolated complementarity determining regions (CDRs), and (7) single chain Fvs (scFvs), the latter of which are preferred (e.g., from an scFV library). Examples of embodiments of the antigen-binding molecules according to the present invention are described, for example, in WO 00 / 006605, WO 2005 / 040220, WO 2008 / 119567, WO 2010 / 037838, WO 2013 / 026837, WO 2013 / 026833, U.S. Patent Application Publication Nos. 2014 / 0308285, 2014 / 0302037, WO 2014 / 144722, WO 2014 / 151910, and WO 2015 / 048272.

[0058] The definition of "binding domain" or "domain that binds to" also includes fragments of full-length antibodies, such as VH, VHH, VL, (s)dAb, Fv, Fd, Fab, Fab', F(ab')2, or "r IgG" ("half antibodies"). The antigen-binding molecules of the present invention may also include modified fragments of antibodies, also called antibody variants, such as scFv, di-scFv or bi(s)-scFv, scFv-Fc, scFv-zipper, scFab, Fab2, Fab3, diabodies, single chain diabodies, tandem diabodies (Tandab's), tandem di-scFv, tandem tri-scFv, "multibodies" such as triabodies or tetrabodies, and single domain antibodies, such as nanobodies, or single variable domain antibodies, comprising only one variable domain, which may be VHH, VH or VL, that specifically binds to an antigen or epitope independently of other V regions or domains.

[0059] As used herein, the term "single-chain Fv", "single-chain antibody", or "scFv" refers to a single polypeptide chain antibody fragment that contains the variable regions from both the heavy and light chains but lacks the constant region. Typically, single-chain antibodies further comprise a polypeptide linker between the VH and VL domains that allows them to form the desired structure that allows for antigen binding. Single-chain antibodies are discussed in detail by Pluckthun in The Pharmacology of Monoclonal Antibodies, vol. 113, Rosenburg and Moore eds. Springer-Verlag, New York, pp. 269-315 (1994). Various methods of making single chain antibodies are known and include those described in U.S. Patent Nos. 4,694,778 and 5,260,203; International Patent Publication WO 88 / 01649; Bird (1988) Science 242:423-442; Huston et al. (1988) Proc. Natl. Acad. Sci. USA 85:5879-5883; Ward et al. (1989) Nature 334:54454; Skerra et al. (1988) Science 242:1038-1041. In certain embodiments, single chain antibodies may be bispecific, multispecific, human, and / or humanized, and / or synthetic.

[0060] Furthermore, the definition of the term "antigen-binding molecule" preferably includes polyvalent / multivalent constructs and therefore bispecific molecules (bispecific means specifically binding to two cell types (i.e. target cells and effector cells) that contain different antigenic structures). The antigen-binding molecules of the present invention are polyvalent / multivalent molecules that specifically bind to three or more (preferably three) antigenic structures via different binding domains in the context of the present invention, which are typically two target binding domains and one CD3 binding domain, preferably CD20 and CD22 targeting. Furthermore, the definition of the term "antigen-binding molecule" includes molecules that consist of only one peptide chain, and molecules that consist of multiple polypeptide chains, which chains may be identical (homodimers, homotrimers, or homooligomers) or different (heterodimers, heterotrimers, or heterooligomers). Such molecules comprising multiple polypeptide chains (i.e. typically two chains) have these chains typically linked to each other as heterodimers by charge pair bonds, e.g. in hetero-Fc entities that function as half-life extenders, e.g. at the C-terminal position of the CD3-binding moieties described herein. Examples of the above identified antigen-binding molecules, e.g. antibody-based molecules, are described, inter alia, in Harlow and Lane, Antibodies a laboratory manual, CSHL Press (1988) and Using Antibodies: a laboratory manual, CSHL Press (1999), Kontermann and Duebel, Antibody Engineering, Springer, 2nd ed. 2010 and Little, Recombinant Antibodies for Immunotherapy, Cambridge University Press 2009.

[0061] The term "bispecific" as used herein refers to an antigen-binding molecule that is "at least bispecific", i.e., that addresses two different cell types, i.e., targets effector cells, and comprises at least a first binding domain and a second binding domain, where at least one binding domain preferably binds to an antigen or target selected from CS1, BCMA, CD20, CD22, FLT3, CD123, MSLN, CLL1, and EpCAM, and another binding domain of the same molecule binds to another antigen or target (herein CD3). Thus, an antigen-binding molecule according to the present invention comprises specificity for at least two different antigens or targets. For example, one domain preferably does not bind to one or more extracellular epitopes of CD3e of the species described herein.

[0062] The term "target cell surface antigen" refers to an antigenic structure present on the cell surface such that it is expressed by the cell and is accessible to the antigen binding molecules described herein. A preferred target cell surface antigen in the context of the present invention is a tumor-associated antigen (TAA). It may be a protein (preferably the extracellular portion of a protein), or a carbohydrate structure (preferably a carbohydrate structure of a protein, such as a glycoprotein). It is preferably a tumor antigen. The term "bispecific antigen binding molecule" of the present invention also encompasses multispecific antigen binding molecules, such as trispecific antigen binding molecules comprising three binding domains, or constructs with more than three (e.g., four, five...) specificities.

[0063] Preferred in the context of the present invention are molecules that are "multispecific", which is understood herein to be "at least bispecific". In this regard, multispecific molecules, such as antigen-binding molecules, are specific for an effector, such as CD3 (more preferably CD3e), and for at least two target cell surface antigens. Said specificity is conferred by the respective binding domains as defined herein. Typically, "multispecific" refers to a molecule that is specific for two different target cell surface effectors, and thus multispecificity confers favorable properties of the multispecific antigen-binding molecules of the present invention (i.e., reduced antigen deficiency and increased therapeutic breadth or better tolerability).

[0064] When the antigen-binding molecule of the present invention is (at least) bispecific, it does not occur in nature and it is significantly different from naturally occurring products. Thus, a "bispecific" antigen-binding molecule or immunoglobulin is an artificial hybrid antibody or immunoglobulin having at least two different binding sites with different specificities. Bispecific antigen-binding molecules can be produced by various methods including fusion of hybridomas or linking of Fab' fragments. See, for example, Songsivilai & Lachmann, Clin. Exp. Immunol. 79:315-321 (1990).

[0065] At least three binding domains and a variable domain (VH / VL) of the antigen-binding molecule of the present invention typically comprise a peptide linker (spacer peptide). The term "peptide linker" according to the present invention includes an amino acid sequence that links the amino acid sequences of one (variable and / or binding) domain and the other (variable and / or binding) domain of the antigen-binding molecule of the present invention to each other. The peptide linker between the first and second binding domains, which can simultaneously bind to two targets, preferably different targets (e.g., TAA1 and TAA2), is preferably flexible and of limited length (e.g., 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, or 18 amino acids). This peptide linker may also be used to fuse the third domain to other domains of the antigen-binding molecule of the present invention. An essential technical feature of such a peptide linker is that it does not contain polymerization activity. Suitable peptide linkers are those described in US Pat. Nos. 4,751,180 and 4,935,233, or WO 88 / 09344. Peptide linkers may also be used to link other domains or modules or regions (such as half-life extension domains) to the antigen-binding molecules of the present invention. Typically, however, the linker between the first and second target-binding domains is different from the intraconjugate linker linking the VH and VL in the target-binding domain. The difference is that the linker between the first and second binding domains has one more amino acid than the intraconjugate linker (e.g., SGGGGS vs. GGGGS, 6 and 5 amino acids, respectively). This surprisingly confers flexibility and stability at the same time in certain antigen-binding molecule formats described herein.

[0066] The antigen-binding molecule of the present invention is preferably an "in vitro generated antigen-binding molecule". This term refers to an antigen-binding molecule as defined above, in which all or a part of the variable region (e.g., at least one CDR) is generated in the selection of non-immune cells, such as in vitro phage display, protein chips, or any other method that can test candidate sequences for antigen-binding ability. Thus, this term preferably excludes sequences that are generated only by genome rearrangement in immune cells of animals. A "recombinant antibody" is an antibody produced by the use of recombinant DNA technology or genetic engineering.

[0067] The term "monoclonal antibody" (mAb) or monoclonal antigen-binding molecule, as used herein, refers to an antibody obtained from a population of substantially homogeneous antibodies, i.e., individual antibodies comprising the population that are identical except for possible naturally occurring mutations and / or post-translational modifications (e.g., isomerization, amidation) that may be present in minor amounts. Monoclonal antibodies are highly specific and are directed against a single antigenic site or determinant on the antigen, in contrast to conventional (polyclonal) antibody preparations that typically contain different antibodies directed against different determinants (or epitopes). In addition to their specificity, monoclonal antibodies are advantageous in that they are synthesized by hybridoma culture and are therefore uncontaminated by other immunoglobulins. The modifier "monoclonal" indicates the character of the antibody as being obtained from a substantially homogeneous population of antibodies and is not to be construed as requiring production of the antibody by any particular method.

[0068] For the preparation of monoclonal antibodies, any technique that results in antibodies produced by continuous cell line culture may be used. For example, the monoclonal antibodies used may be produced by the hybridoma method first described by Koehler et al., Nature, 256:495 (1975), or may be produced by recombinant DNA methods (see, for example, U.S. Pat. No. 4,816,567). Additional examples of techniques for producing human monoclonal antibodies include the trioma technique, the human B-cell hybridoma technique (Kozbor, Immunology Today 4 (1983), 72), and the EBV-hybridoma technique (Cole et al., Monoclonal Antibodies and Cancer Therapy, Alan R. Liss, Inc. (1985), 77-96).

[0069] The hybridomas can then be screened using standard methods such as enzyme-linked immunosorbent assay (ELISA) and surface plasmon resonance analysis, e.g., Biacore™, to identify one or more hybridomas that produce antibodies that specifically bind to the designated antigen. Any form of the relevant antigen can be used as an immunogen, for example, recombinant antigen, naturally occurring form, any variant or fragment thereof, and antigenic peptides thereof. Surface plasmon resonance, as employed in the Biacore system, can be used to increase the efficiency of binding of phage antibodies to epitopes of surface antigens of target cells (Schier, Human Antibodies Hybridomas 7 (1996), 97-105; Malmborg, J. Immunol. Methods 183 (1995), 7-13).

[0070] Another exemplary method for generating monoclonal antibodies includes screening protein expression libraries, such as phage display or ribosome display libraries. Phage display is described, for example, in Ladner et al., U.S. Pat. No. 5,223,409, Smith (1985) Science 228:1315-1317, Clackson et al., Nature, 352:624-628 (1991), and Marks et al., J. Mol. Biol., 222:581-597 (1991).

[0071] In addition to using display libraries, the relevant antigens can be used to immunize non-human animals, such as rodents (e.g., mice, hamsters, rabbits, or rats). In one embodiment, the non-human animals contain at least a portion of a human immunoglobulin gene. For example, mouse strains deficient in mouse antibody production can be engineered with large fragments of the human Ig (immunoglobulin) locus. Hybridoma technology can be used to produce and select antigen-specific monoclonal antibodies derived from genes with the desired specificity. See, e.g., XENOMOUSE™, Green et al. (1994) Nature Genetics 7:13-21, U.S. Patent Application Publication No. 2003-0070185, WO 96 / 34096, and WO 96 / 33735.

[0072] Monoclonal antibodies may also be obtained from non-human animals and then modified, for example, humanized, deimmunized, chimerized, etc., using recombinant DNA techniques known in the art. Examples of modified antigen-binding molecules include humanized variants of non-human antibodies, "affinity matured" antibodies (see, for example, Hawkins et al. J. Mol. Biol. 254, 889-896 (1992) and Lowman et al., Biochemistry 30, 10832-10837 (1991)), and antibody variants with modified effector functions (see, for example, U.S. Pat. No. 5,648,260, Kontermann and Duebel (2010), supra, and Little (2009), supra).

[0073] In immunology, affinity maturation is the process by which B cells produce antibodies with increased affinity for antigens during an immune response. Repeated exposure to the same antigen causes the host to produce antibodies with successively higher affinities. Similar to natural prototypes, in vitro affinity maturation is based on the principle of mutation and selection. In vitro affinity maturation has been successfully used to optimize antibodies, antigen-binding molecules, and antibody fragments. Random mutations within the CDRs are introduced using radiation, chemical mutagens, or error-prone PCR. In addition, chain shuffling can increase genetic diversity. Two or three rounds of mutation and selection using display methods such as phage display usually result in antibody fragments with affinities in the low nanomolar range.

[0074] A preferred type of amino acid substitution variant of an antigen-binding molecule involves the substitution of one or more hypervariable region residues of a parent antibody (e.g., a humanized or human antibody). Generally, the resulting variants selected for further development will have improved biological properties relative to the parent antibody from which they were generated. A convenient method for generating such substitution variants involves affinity maturation using phage display. Briefly, several hypervariable region sites (e.g., 6-7 sites) are mutated to generate all possible amino acid substitutions at each site. The antibody variants so generated are displayed in a monovalent form from filamentous phage particles as fusions with the gene III product of M13 packaged within each particle. The phage-displayed variants are then screened for their biological activity (e.g., binding affinity) as disclosed herein. To identify candidate hypervariable region sites for modification, alanine scanning mutagenesis can be performed to identify hypervariable region residues that contribute significantly to antigen binding. Alternatively or additionally, it may be beneficial to analyze a crystal structure of the antigen-antibody complex to identify contact points between the binding domain and, for example, human CS1, BCMA, CD20, CD22, FLT3, CD123, CDH3, MSLN, CLL1, or EpCAM. Such contact and adjacent residues are candidates for substitution according to the techniques detailed herein. After generating such variants, the panel of variants may be subjected to screening as described herein, and antibodies with superior properties in one or more relevant assays may be selected for further development.

[0075] The monoclonal antibodies and antigen-binding molecules of the present invention include, inter alia, "chimeric" antibodies (immunoglobulins) in which a portion of the heavy and / or light chain is identical or homologous to corresponding sequences in antibodies derived from a particular species or belonging to a particular antibody class or subclass, while the remainder of the chain is identical or homologous to corresponding sequences in antibodies derived from another species or belonging to another antibody class or subclass, as well as fragments of such antibodies, so long as they exhibit the desired biological activity (U.S. Pat. No. 4,816,567; Morrison et al., Proc. Natl. Acad. Sci. USA, 81:6851-6855 (1984)). Chimeric antibodies of interest herein include "primatized" antibodies that contain variable domain antigen-binding sequences derived from a non-human primate (e.g., Old World monkey, ape, etc.) and human constant region sequences. Various methods for producing chimeric antibodies have been described. See, e.g., Morrison et al., Proc. Natl. Acad. ScL USA 81:6851, 1985; Takeda et al., Nature 314:452, 1985; Cabilly et al., U.S. Pat. No. 4,816,567; Boss et al., U.S. Pat. No. 4,816,397; Tanaguchi et al., European Patent Nos. 0171496; 0173494; and British Patent No. 2177096.

[0076] Antibodies, antigen-binding molecules, antibody fragments, or antibody variants may also be modified by specific deletion of human T-cell epitopes (a method called "deimmunization"), for example by the methods disclosed in WO 98 / 52976 or WO 00 / 34317. Briefly, the heavy and light chain variable domains of an antibody may be analyzed for peptides that bind to MHC class II. These peptides represent potential T-cell epitopes (as defined in WO 98 / 52976 and WO 00 / 34317). For the detection of potential T-cell epitopes, a computer modeling method called "peptide threading" may be applied, as described in WO 98 / 52976 and WO 00 / 34317, and in addition, a database of human MHC class II binding peptides may be searched for motifs present in the VH and VL sequences. These motifs bind to any of the 18 major MHC class II DR allotypes and therefore constitute potential T cell epitopes. Detected potential T cell epitopes can be eliminated by substituting a small number of amino acid residues in the variable domains, or preferably by single amino acid substitutions. Typically, conservative substitutions are made. Often, but not exclusively, amino acids common to positions in human germline antibody sequences can be used. Human germline sequences are disclosed, for example, in Tomlinson, et al. (1992) J. Mol. Biol. 227:776-798; Cook, GP et al. (1995) Immunol. Today Vol. 16(5):237-242; and Tomlinson et al. (1995) EMBO J. 14:14:4628-4638. The V BASE directory provides a comprehensive directory of human immunoglobulin variable region sequences (compiled by Tomlinson, LA. et al. MRC Centre for Protein Engineering, Cambridge, UK). This sequence can be used as a source of human sequences, for example, for the framework regions and CDRs. For example, the consensus human framework regions described in U.S. Patent No. 6,300,064 can also be used.

[0077] A "humanized" antibody, antigen-binding molecule, variant, or fragment thereof (e.g., Fv, Fab, Fab', F(ab')2, or other antigen-binding subsequence of an antibody) is an antibody or immunoglobulin of mostly human sequence that contains minimal sequence derived from non-human immunoglobulin. In most cases, humanized antibodies are human immunoglobulins (recipient antibody) in which residues from a hypervariable region (also referred to as CDR) of the recipient are replaced by residues from a hypervariable region of a non-human (e.g., rodent) species (donor antibody) such as mouse, rat, hamster, or rabbit having the desired specificity, affinity, and capacity. In some cases, Fv framework region (FR) residues of the human immunoglobulin are replaced by corresponding non-human residues. Furthermore, "humanized antibodies", as the term is used herein, may also comprise residues that are not found in the recipient antibody or in the donor antibody. These modifications are made to further refine and optimize antibody performance. A humanized antibody may also comprise at least a portion of an immunoglobulin constant region (Fc), typically that of a human immunoglobulin. For further details, see Jones et al., Nature, 321:522-525 (1986); Reichmann et al., Nature, 332:323-329 (1988); and Presta, Curr. Op. Struct. Biol., 2:593-596 (1992).

[0078] Humanized antibodies or fragments thereof can be generated by replacing sequences of Fv variable domains not directly involved in antigen binding with equivalent sequences from human Fv variable domains. Exemplary methods for generating humanized antibodies or fragments thereof are provided by Morrison (1985) Science 229:1202-1207; Oi et al. (1986) BioTechniques 4:214 and U.S. Patent Nos. 5,585,089, 5,693,761, 5,693,762, 5,859,205, and 6,407,213. These methods include isolating, manipulating, and expressing nucleic acid sequences that encode all or part of an immunoglobulin Fv variable domain from at least one of the heavy or light chains. Such nucleic acids may be obtained from hybridomas producing antibodies against a predetermined target as described above, as well as other sources. The recombinant DNA encoding the humanized antibody molecule can then be cloned into an appropriate expression vector.

[0079] Humanized antibodies may also be produced using transgenic animals, such as mice, that express human heavy and light chain genes but are incapable of expressing endogenous mouse immunoglobulin heavy and light chain genes. Winter describes an exemplary CDR grafting method that may be used to prepare the humanized antibodies described herein (U.S. Pat. No. 5,225,539). All of the CDRs of a particular human antibody may be replaced with at least a portion of a non-human CDR, or only a portion of the CDRs may be replaced with a non-human CDR. It is only necessary to replace as many CDRs as are required for the binding of the humanized antibody to a given antigen.

[0080] Humanized antibodies may be optimized by introducing conservative substitutions, consensus sequence substitutions, germline substitutions, and / or back mutations. Such modified immunoglobulin molecules may be generated by any of several techniques known in the art (e.g., Teng et al., Proc. Natl. Acad. Sci. USA, 80:7308-7312, 1983; Kozbor et al., Immunology Today, 4:7279, 1983; Olsson et al., Meth. Enzymol., 92:3-16, 1982, and EP 239400).

[0081] The terms "human antibody," "human antigen-binding molecule," and "human binding domain" include antibodies, antibody-binding molecules, and binding domains having antibody regions, such as variable and constant regions or domains, that substantially correspond to human germline immunoglobulin sequences known in the art, including, for example, those described in Kabat et al. (1991) supra. The human antibodies, antigen-binding molecules, or binding domains of the invention may contain amino acid residues not encoded by human germline immunoglobulin sequences (e.g., mutations introduced by random or site-specific mutagenesis in vitro or by somatic mutation in vivo), for example in the CDRs, particularly CDR3. The human antibodies, antigen-binding molecules, or binding domains may have at least one, two, three, four, five, or more positions replaced with amino acid residues not encoded by human germline immunoglobulin sequences. The definitions of human antibody, antigen-binding molecule, and binding domain as used herein also contemplate fully human antibodies that contain only human sequences of non-artificially and / or genetically modified antibodies, which can be obtained by using technologies or systems such as Xenomouse. Preferably, a "fully human antibody" does not contain any amino acid residues that are not encoded by human germline immunoglobulin sequences.

[0082] In some embodiments, the antigen-binding molecules of the present invention are "isolated" or "substantially pure" antigen-binding molecules. "Isolated" or "substantially pure", when used to describe the antigen-binding molecules disclosed herein, refers to an antigen-binding molecule that has been identified, separated, and / or recovered from components of its production environment. Preferably, the antigen-binding molecule is free or substantially free of association with all other components from its production environment. Contaminating components of its production environment, such as components resulting from recombinant transfected cells, are materials that would normally interfere with diagnostic or therapeutic uses for the polypeptide, which may include enzymes, hormones, and other proteinaceous or non-proteinaceous solutes. The antigen-binding molecule may, for example, constitute at least about 5% by weight, or at least about 50% by weight, of the total protein in a given sample. It is understood that an isolated protein may constitute 5% to 99.9% by weight of the total protein content, depending on the circumstances. The polypeptide may be produced at significantly higher concentrations by using an inducible promoter or a high expression promoter, resulting in increased concentration levels. This definition includes the production of antigen-binding molecules in a wide variety of organisms and / or host cells known in the art. In a preferred embodiment, the antigen-binding molecule will be purified (1) to a sufficient extent to obtain at least 15 residues of the N-terminal or internal amino acid sequence by using a spinning cup sequenator, or (2) to homogeneity by SDS-PAGE under non-reducing or reducing conditions using Coomassie blue or preferably silver staining. However, typically, an isolated antigen-binding molecule will be prepared by at least one purification step.

[0083] The term "binding domain" in the context of the present invention is characterized by a domain that (specifically) binds to / interacts with / recognizes a given target epitope or a given target site on a target molecule (antigen), such as CD20 and CD22 and CD3, respectively. The structure and function of the first and / or second binding domain (recognizing CD20 and CD22), and preferably also the structure and / or function of the effector binding domain (typically the third binding domain recognizing CD3), are based on the structure and / or function of an antibody (e.g. a full-length or complete immunoglobulin molecule) and / or are derived from the variable heavy (VH) and / or variable light (VL) domains of an antibody or a fragment thereof. Preferably, the target cell surface antigen binding domain is characterized by the presence of three light chain CDRs (i.e. CDR1, CDR2 and CDR3 of the VL region) and / or three heavy chain CDRs (i.e. CDR1, CDR2 and CDR3 of the VH region). The effector (typically CD3) binding domain also preferably comprises the minimum structural requirements of an antibody to enable target binding. More preferably, the second binding domain comprises at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 of the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 of the VH region). It is envisaged that the first binding domain and / or the second binding domain are generated or obtained by phage display or library screening methods other than grafting CDR sequences from an existing (monoclonal) antibody onto the scaffold.

[0084] According to the present invention, the binding domain is in the form of one or more polypeptides. Such polypeptides may contain proteinaceous and non-proteinaceous parts (e.g. chemical linkers or chemical cross-linking agents such as glutaraldehyde). Proteins (including fragments thereof, preferably biologically active fragments, and peptides, usually having less than 30 amino acids) comprise two or more amino acids coupled to each other via covalent peptide bonds (resulting in a chain of amino acids).

[0085] The term "polypeptide" as used herein refers to a group of molecules that usually consists of more than 30 amino acids. Polypeptides may further form multimers, such as dimers, trimers, and higher oligomers, i.e., they may consist of multiple polypeptide molecules. The polypeptide molecules forming such dimers, trimers, etc. may be identical or non-identical. The corresponding higher order structures of such multimers are therefore referred to as homo- or heterodimers, homo- or heterotrimers, etc. An example of a heteromultimer is an antibody molecule, which in its native form consists of two identical polypeptide light chains and two identical polypeptide heavy chains. The terms "peptide", "polypeptide", and "protein" also refer to naturally occurring modified peptides / polypeptides / proteins that have been modified, for example, by post-translational modifications such as glycosylation, acetylation, phosphorylation, etc. A "peptide", "polypeptide", or "protein", as referred to herein, may also be chemically modified, such as pegylation. Such modifications are known in the art and are described herein below.

[0086] Preferably, the binding domain that binds to any of CS1, BCMA, CD20, CD22, FLT3, CD123, CLL1, CDH3, MSLN, and EpCAM, and / or the binding domain that binds to CD3ε, is a human binding domain. Antibodies and antigen-binding molecules that include at least one human binding domain avoid some of the problems associated with antibodies or antigen-binding molecules that have non-human variable and / or constant regions, such as from rodents (e.g., mice, rats, hamsters, or rabbits). The presence of such rodent-derived proteins may result in rapid clearance of the antibody or antigen-binding molecule, or may generate an immune response against the antibody or antigen-binding molecule by the patient. To avoid the use of rodent-derived antibodies or antigen-binding molecules, human or fully human antibodies / antigen-binding molecules may be generated by introducing human antibody functions into rodents so that the rodents produce fully human antibodies.

[0087] The ability to clone and reconstruct megabase-sized human loci in yeast artificial chromosomes (YACs) and introduce them into the mouse germline provides a powerful approach for elucidating the functional elements of very large or coarsely mapped loci and for generating useful models of human disease. Furthermore, the use of such techniques to replace mouse loci with their human equivalents will provide unique insight into the expression and regulation of nascent human gene products, their transfer to other systems, and their involvement in disease induction and progression.

[0088] An important practical application of such a strategy is the "humanization" of the mouse humoral immune system. The introduction of human immunoglobulin (Ig) loci into mice in which the endogenous Ig genes have been inactivated provides an opportunity to study the mechanisms underlying the programmed expression and assembly of antibodies and their role in B-cell development. Furthermore, such a strategy would provide an ideal source for the generation of fully human monoclonal antibodies (mAbs), which would be a major milestone in realizing the potential of antibody therapy in human diseases. Fully human antibodies or antigen-binding molecules are expected to minimize the immunogenic and allergic reactions inherent to mouse or mouse-derived mAbs, thereby increasing the efficacy and safety of the administered antibodies / antigen-binding molecules. The use of fully human antibodies or antigen-binding molecules can be expected to provide significant advantages in the treatment of chronic and recurrent human diseases that require repeated administration of compounds, such as inflammation, autoimmunity, and cancer.

[0089] One approach towards this goal has been to engineer mouse strains deficient in mouse antibody production with large fragments of the human Ig loci, with the expectation that such mice would produce a broad repertoire of human antibodies without mouse antibodies. The large human Ig fragments would retain the broad diversity of variable genes and the proper regulation of antibody production and expression. By utilizing the mouse machinery for antibody diversification and selection, as well as the lack of immune tolerance to human proteins, the human antibody repertoires recapitulated in these mouse strains should produce high affinity antibodies against any antigen of interest, including human antigens. Using hybridoma technology, antigen-specific human mAbs with the desired specificity could be readily generated and selected. This general strategy was demonstrated in conjunction with the generation of the first XenoMouse mouse strains (see Green et al. Nature Genetics 7:13-21 (1994)). This XenoMouse strain was engineered with YACs containing 245 kb and 190 kb germline-configured fragments of the human heavy and kappa light chain loci, respectively, that contained the core sequences of the variable and constant regions. The human Ig-containing YACs proved compatible with the mouse system for both antibody rearrangement and expression, and were able to replace inactivated mouse Ig genes. This was demonstrated by their ability to induce B cell development to produce an adult-like human repertoire of fully human antibodies and to produce antigen-specific human mAbs. These results also suggested that the introduction of a large portion of the human Ig locus, containing multiple V genes, additional regulatory elements, and human Ig constant regions, could recapitulate a virtually complete repertoire characterized by the human humoral response to infection and immunization. Recently, extending the work of Green et al., the introduction of megabase-sized germline-configured YAC fragments of the human heavy and kappa light chain loci, respectively, introduced approximately 80% of the human antibody repertoire. See Mendez et al. Nature Genetics 15:146-156 (1997) and U.S. patent application Ser. No. 08 / 759,620.

[0090] The generation of XenoMouse animals is disclosed in U.S. Patent Application Serial Nos. 07 / 466,008, 07 / 610,515, 07 / 919,297, 07 / 922,649, 08 / 031,801, 08 / 112,848, 08 / 234,145, 08 / 376,279, 08 / 430,938, 08 / 464,584, 08 / 464,582, 08 / 463,191, 08 / 462,837, 08 / 486, 853, 08 / 486,857, 08 / 486,859, 08 / 462,513, 08 / 724,752, and 08 / 759,620; and U.S. Pat. Nos. 6,162,963; 6,150,584; 6,114,598; 6,075,181, and 5,939,598; and Japanese Patent Publications Nos. 3068180B2, 3068506B2, and 3068507B2. See also Mendez et al. Nature Genetics 15:146-156 (1997) and Green and Jakobovits J. Exp. Med. 188:483-495 (1998), EP 0463151 B1, WO 94 / 02602, WO 96 / 34096, WO 98 / 24893, WO 00 / 76310, and WO 03 / 47336.

[0091] In another approach, others including GenPharm International, Inc. have utilized a "minilocus" approach. In the minilocus method, an exogenous Ig locus is mimicked through the inclusion of pieces (individual genes) from the Ig locus. Thus, one or more VH genes, one or more DH genes, one or more JH genes, a μ constant region, and a second constant region (preferably a γ constant region) are formed as a construct for insertion into an animal. This approach is described in U.S. Pat. No. 5,545,807 to Surani et al., and U.S. Pat. Nos. 5,545,806, 5,625,825, 5,625,126, 5,633,425, 5,661,016, 5,770,429, 5,789,650, 5,814,318, and 5,877,397, all to Lonberg and Kay, respectively. Nos. 5,874,299, and 6,255,458 to Krimpenfort and Berns, U.S. Pat. Nos. 5,591,669 and 6,023.010 to Krimpenfort and Berns, U.S. Pat. Nos. 5,612,205, 5,721,367, and 5,789,215 to Berns et al., and U.S. Pat. No. 5,643,763 to Choi and Dunn, and GenPharm and U.S. Patent Application Nos. 07 / 574,748, 07 / 575,962, 07 / 810,279, 07 / 853,408, 07 / 904,068, 07 / 990,860, 08 / 053,131, 08 / 096,762, 08 / 155,301, 08 / 161,739, 08 / 165,699, and 08 / 209,741 to International, all of which are incorporated herein by reference.See also EP 0546073B1, WO 92 / 03918, WO 92 / 22645, WO 92 / 22647, WO 92 / 22670, WO 93 / 12227, WO 94 / 00569, WO 94 / 25585, WO 96 / 14436, WO 97 / 13852, and WO 98 / 24884, and U.S. Pat. No. 5,981,175. See also Taylor et al. (1992), Chen et al. (1993), Tuaillon et al. (1993), Choi et al. (1993), Lonberg et al. (1994), Taylor et al. (1994), and Tuaillon et al. (1995), Fishwild et al. (1996).

[0092] Kirin has also demonstrated the production of human antibodies from mice into which large chromosome pieces or entire chromosomes have been introduced by microcell fusion. See EP 773288 and EP 843961. Xenerex Biosciences is developing a promising human antibody production technology in which SCID mice are reconstituted with human lymphoid cells, e.g., B and / or T cells. The mice can then be immunized with an antigen to generate an immune response against the antigen. See U.S. Pat. Nos. 5,476,996, 5,698,767, and 5,958,765.

[0093] Human anti-mouse antibody (HAMA) reactions have led the industry to produce chimeric or otherwise humanized antibodies. However, it is expected that some human anti-chimeric antibody (HACA) reactions will be observed, especially in chronic or multiple dose applications of the antibodies. Therefore, it would be desirable to provide an antigen-binding molecule that includes a human binding domain for CS1, BCMA, CD20, CD22, FLT3, CD123, CLL1, MSLN, CDH3, or EpCAM and a human binding domain for CD3ε to eliminate the concerns and / or impact of HAMA or HACA reactions.

[0094] The terms "(specifically) or (immunospecifically) bind", "(specifically) recognize", "(specifically) induce" and "(specifically) react" mean according to the present invention that a binding domain, preferably by its paratope, interacts or specifically interacts with a given epitope or a given target site on a target molecule (antigen, here preferably CS1, BCMA, CD20, CD22, FLT3, CD123, CLL1, MSLN, CDH3, or EpCAM, and CD3ε, respectively).

[0095] In the context of the present invention, a paratope is understood as an antigen-binding site that is part of a polypeptide as described herein and that recognizes and binds to an antigen. A paratope is typically a small region of at least about 5 amino acids. A paratope as understood herein typically comprises a portion of the heavy (VH) and light (VL) chain sequences derived from an antibody. Each binding domain of the polypeptide according to the present invention comprises a paratope that comprises a set of six complementarity determining regions (CDR loops), three of which are contained within the VH and VL sequences derived from an antibody.

[0096] In the context of the present invention, the antigen-binding molecules (i.e., preferably polypeptides) of the present invention bind to their respective target structures in a specific manner. Preferably, the polypeptides according to the present invention comprise one paratope per binding domain that "specifically or immunospecifically binds" to, "(specifically or immunospecifically) recognizes" or "(specifically or immunospecifically) reacts" with the respective target structure. This means that, according to the present invention, the polypeptide or its binding domain interacts or (immuno)specifically interacts with a given epitope on the target molecule (antigen) and CD3, respectively. This interaction or association occurs more frequently, more rapidly, more persistently, with greater affinity, or some combination of these parameters, for an epitope on a particular target compared to an alternative substance (non-target molecule). However, due to sequence similarity between homologous proteins in various species, an antibody construct or binding domain that immunospecifically binds to a target (e.g., a human target) may cross-react with a homologous target molecule from a different species (e.g., a non-human primate). Thus, the term "specific / immunospecific binding" can include binding of an antibody construct or binding domain to an epitope in multiple species and / or to a structurally related epitope. The term "(immuno)selectively binds" excludes binding to a structurally related epitope.

[0097] The term "epitope" refers to a site on an antigen to which a binding domain, such as an antibody or immunoglobulin, or a derivative, fragment, or variant of an antibody or immunoglobulin, specifically binds. An "epitope" is antigenic, and therefore the term epitope is also sometimes referred to herein as an "antigenic structure" or "antigenic determinant". Thus, the binding domain is an "antigen interaction site". Said binding / interaction is also understood to define "specific recognition".

[0098] An "epitope" can be formed by both contiguous amino acids or non-contiguous amino acids juxtaposed by tertiary folding of a protein. A "linear epitope" is an epitope that comprises an epitope in which a primary amino acid sequence is recognized. A linear epitope typically comprises at least 3 or at least 4, and more usually at least 5, or at least 6, or at least 7 (e.g., about 8 to about 10) amino acids in a unique sequence.

[0099] A "conformational epitope", in contrast to a linear epitope, is an epitope in which the primary sequence of amino acids that comprise the epitope is not the only element that defines the recognized epitope (e.g., an epitope in which the primary sequence of amino acids is not necessarily recognized by the binding domain). Typically, a conformational epitope comprises a larger number of amino acids compared to a linear epitope. With respect to the recognition of a conformational epitope, the binding domain recognizes the three-dimensional structure of the antigen (preferably a peptide or protein or a fragment thereof) (in the context of the present invention, the antigenic structure for one of the binding domains is contained within the surface antigen protein of the target cell). For example, when a protein molecule folds to form a three-dimensional structure, certain amino acids and / or polypeptide backbones that form a conformational epitope are juxtaposed, thereby allowing the antibody to recognize the epitope. Methods for determining the conformational structure of an epitope include, but are not limited to, x-ray crystallography, two-dimensional nuclear magnetic resonance (2D-NMR) spectroscopy, and site-specific spin labeling and electron paramagnetic resonance (EPR) spectroscopy.

[0100] Methods for epitope mapping are described below: if a region (a contiguous stretch of amino acids) of human CD20 and CD22 proteins is exchanged or replaced with the corresponding region of non-human and non-primate CD20 and CD22 (e.g. mouse CD20 and CD22, but also chicken, rat, hamster, rabbit, etc.), a reduction in the binding activity of the binding domain is expected to occur, as long as the binding domain is not cross-reactive with the non-human, non-primate CD20 and CD22 used. Said reduction is preferably at least 10%, 20%, 30%, 40%, or 50%; more preferably at least 60%, 70%, or 80%, and most preferably 90%, 95%, or even 100% compared to the binding to the region in human CD20 and CD22 CD20 and CD22 proteins, respectively, when the binding to the corresponding region of human CD20 and CD22 proteins is taken as 100%. It is envisaged that the above human CD20 and CD22 / non-human CD20 and CD22 chimeras are expressed in CHO cells. It is also envisaged that the human CD20 and CD22 / non-human CD20 and CD22 chimeras are fused to the transmembrane and / or cytoplasmic domains of different membrane-associated proteins, such as EpCAM.

[0101] In an alternative or additional method of epitope mapping, several truncations of human CD20 and CD22 extracellular domains can be made to determine the specific regions recognized by the binding domains. In these truncations, different extracellular CD20 and CD22 domains / subdomains or regions are deleted stepwise starting from the N-terminus. It is envisaged that the truncated CD20 and CD22 can be expressed in CHO cells. It is also envisaged that the truncated CD20 and CD22 may be fused to the transmembrane and / or cytoplasmic domains of different membrane-bound proteins such as EpCAM. It is also envisaged that the truncated CD20 and CD22 may include a signal peptide domain at their N-terminus (e.g., a signal peptide derived from the mouse IgG heavy chain signal peptide). It is further envisaged that the truncated CD20 and CD22 may include a v5 domain at the N-terminus (following the signal peptide) that can confirm their correct expression on the cell surface. Truncated forms of CD20 and CD22 that no longer encompass the region of CD20 and CD22 recognized by the binding domains are expected to result in reduced or lost binding, preferably at least 10%, 20%, 30%, 40%, or 50%; more preferably at least 60%, 70%, 80%, and most preferably 90%, 95%, or even 100%, when binding to the whole human CD20 and CD22 proteins (or their extracellular regions or domains) is taken as 100%.

[0102] A further method for determining the contribution of specific residues of CD20 and CD22 to recognition by antibody binding molecules or binding domains is alanine scanning (see, for example, Morrison KL & Weiss GA. Cur Opin Chem Biol. 2001 Jun;5(3):302-7), in which each residue to be analyzed is replaced with alanine, for example by site-directed mutagenesis. Alanine is used because it is not bulky, is chemically inert, and yet has a methyl functionality that mimics the secondary structure criteria that many other amino acids have. If it is desirable to preserve the size of the residue to be mutated, bulky amino acids such as valine or leucine can sometimes be used. Alanine scanning is a mature technique that has been used for a long time.

[0103] The interaction between a binding domain and an epitope or an epitope-containing region means that the binding domain exhibits a significant affinity for the epitope / epitope-containing region on a particular protein or antigen (herein, CD20 and CD22, and CD3, respectively), and generally does not exhibit significant reactivity with proteins or antigens other than CD20 and CD22 or CD3. "Measurable affinity" refers to a binding domain that exhibits a significant affinity of about 10 -6 M(KD) or stronger. Preferably, the binding affinity is about 10 -12 ~10 -8 M, 10 -12 ~10 -9 M, 10 -12 ~10 -10 M, 10 -11 ~10 -8 M, preferably about 10 -11 ~10 -9M, the binding is considered specific. Whether a binding domain specifically reacts with or binds to a target can be easily tested, inter alia, by comparing the reaction of said binding domain to a target protein or antigen with the reaction of said binding domain to proteins or antigens other than CD20, CD22, or CD3. Preferably, the binding domain of the present invention does not essentially or substantially bind to proteins or antigens other than CD20 and CD22 or CD3 (i.e., the first binding domain cannot bind to proteins other than CD20 and the second binding domain cannot bind to proteins other than CD22). It is an envisaged feature of the antigen binding molecule of the present invention to have superior affinity properties compared to other HLE formats. Such superior affinity consequently suggests an extended half-life in vivo. The longer half-life of the antigen binding molecule of the present invention may reduce the duration and frequency of administration, which typically contributes to improved patient compliance. This is particularly important, since the antigen binding molecule of the present invention is particularly beneficial for highly debilitated or even multi-disease cancer patients.

[0104] The terms "does not essentially / substantially bind" or "cannot bind" mean that the binding domain of the invention does not bind to proteins or antigens other than CD20 and CD22 or CD3, i.e. does not show more than 30%, preferably more than 20%, more preferably more than 10%, particularly preferably more than 9%, 8%, 7%, 6% or 5% reactivity with proteins or antigens other than CD20, CD22 or CD3, when the binding to CD20, CD22 or CD3, respectively, is taken as 100%.

[0105] Specific binding is believed to be mediated by specific motifs within the amino acid sequences of the binding domain and the antigen. Thus, binding occurs as a result of their primary, secondary and / or tertiary structure, as well as secondary modifications of said structure. Specific interaction of the antigen interaction site with its specific antigen may result in simple binding of the interaction site to the antigen. Furthermore, specific interaction of the antigen interaction site with its specific antigen may alternatively or additionally result in the initiation of a signal, for example by inducing a conformational change in the antigen, oligomerization of the antigen, etc.

[0106] The term "variable" refers to that portion of an antibody or immunoglobulin domain (i.e., the "variable domain") that exhibits variability in sequence and is responsible for determining the specificity and binding affinity of a particular antibody. The pairing of a variable heavy chain (VH) and a variable light chain (VL) together forms a single antigen-binding site.

[0107] The variability is not uniformly distributed throughout the variable domain of an antibody, but is concentrated in subdomains of each of the heavy and light chain variable regions. These subdomains are called "hypervariable regions" or "complementarity determining regions" (CDRs). The more conserved (i.e., non-hypervariable) parts of the variable domains are called "framework" regions (FRMs or FRs) and provide a scaffold for the six CDRs in the three-dimensional space that forms the antigen-binding surface. Naturally occurring heavy and light chain variable domains each contain four FRM regions (FR1, FR2, FR3, and FR4) that largely adopt a β-sheet configuration, which are connected by three hypervariable regions that form loop connections and, in some cases, form part of the β-sheet structure. The hypervariable regions of each chain are held together in close proximity by the FRMs and, together with the hypervariable regions of the other chain, contribute to the formation of the antigen-binding site (see Kabat et al., supra).

[0108] The term "CDR" and its plural "CDRs" refer to the complementarity determining regions, three of which constitute the binding properties of the light chain variable region (CDR-L1, CDR-L2, and CDR-L3) and three of which constitute the binding properties of the heavy chain variable region (CDR-H1, CDR-H2, and CDR-H3). The CDRs contain most of the residues responsible for the specific interactions of the antibody with the antigen and thus contribute to the functional activity of the antibody molecule: the CDRs are the primary determinants of antigen specificity.

[0109] The exact definition of the boundaries and lengths of CDRs are subject to various classification and numbering systems. Thus, CDRs may be represented by Kabat, Chothia, contact, or any other boundary definition, including the numbering systems described herein. Although the boundaries are different, each of these systems has some overlap in the parts that constitute the so-called "hypervariable regions" in the variable sequences. Thus, the definitions of CDRs according to these systems may differ in length and in the boundary regions with respect to the adjacent framework regions. See, for example, Kabat (an approach based on interspecies sequence variability), Chothia (an approach based on crystallographic studies of antigen-antibody complexes), and / or MacCallum (Kabat et al., supra; Chothia et al., J. Mol. Biol, 1987, 196:901-917; and MacCallum et al., J. Mol. Biol, 1996, 262:732). Yet another standard for characterizing antigen-binding site is the AbM definition used by Oxford Molecular's AbM antibody modeling software.See, for example, Protein Sequence and Structure Analysis of Antibody Variable Domains.In:Antibody Engineering Lab Manual (Ed.:Duebel, S. and Kontermann, R., Springer-Verlag, Heidelberg).As long as two residue identification techniques define overlapping but not identical regions, they can be combined to define hybrid CDRs.However, numbering according to the so-called Kabat system is preferred.

[0110] Typically, CDRs form loop structures that can be classified as canonical structures. The term "canonical structure" refers to the main chain conformation that the antigen-binding (CDR) loop adopts. Comparative structural studies have found that five of the six antigen-binding loops have a limited repertoire of available conformations. Each canonical structure can be characterized by the torsion angle of the polypeptide backbone. Thus, corresponding loops between antibodies can have very similar three-dimensional structures, despite the high amino acid sequence variability found in the majority of the loops (Chothia and Lesk, J. Mol. Biol., 1987, 196:901; Chothia et al., Nature, 1989, 342:877; Martin and Thornton, J. Mol. Biol, 1996, 263:800). Furthermore, there is a relationship between the adopted loop structure and the amino acid sequence surrounding it. The conformation of a particular canonical class is determined by the length of the loop and the amino acid residues present at key positions within the loop as well as within the conserved framework (i.e., outside the loop), and therefore assignment to a particular canonical class can be made based on the presence of these key amino acid residues.

[0111] The term "canonical structure" may also include considerations regarding the linear sequence of an antibody, e.g., as cataloged by Kabat (Kabat et al., supra). The Kabat numbering scheme is a widely adopted standard for numbering the amino acid residues of antibody variable domains in a consistent manner, and is the preferred scheme applied in the present invention, as mentioned elsewhere herein. Additional structural considerations may also be used to determine the canonical structure of an antibody. For example, differences not fully reflected by the Kabat numbering system may be accounted for by the Chothia et al. numbering system and / or may be revealed by other techniques (e.g., crystallography and two- or three-dimensional computer modeling). Thus, a given antibody sequence may be classified into canonical classes that, among other things, allow the identification of suitable chassis sequences (e.g., based on the desire to include various canonical structures in a library). The Kabat numbering system for antibody amino acid sequences and the structural considerations discussed in Chothia et al. (supra) and their significance for interpreting the canonical aspects of antibody structure are explained in the literature. The subunit structures and three-dimensional configurations of various classes of immunoglobulins are known in the art. For a general overview of antibody structure, see Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory, eds. Harlow et al., 1988.

[0112] The CDR3 of the light chain, and especially the CDR3 of the heavy chain, may constitute the most important determinant in antigen binding within the variable regions of the light and heavy chains.In some antigen-binding molecules, the heavy chain CDR3 is likely to constitute the main contact area between antigen and antibody.An in vitro selection scheme that changes only CDR3 can be used to change the binding properties of the antibody or to determine which residues contribute to antigen binding.Therefore, CDR3 is usually the greatest source of molecular diversity within the antibody binding site.For example, H3 can be as short as 2 amino acid residues or more than 26 amino acids.

[0113] In classical full-length antibodies or immunoglobulins, each light (L) chain is linked to a heavy (H) chain by one covalent disulfide bond, while the two H chains are linked to each other by one or more disulfide bonds depending on the H chain isotype. The CH domain closest to the VH is usually referred to as CH1. The constant ("C") domains are not directly involved in antigen binding, but exhibit various effector functions such as antibody-dependent, cell-mediated cytotoxicity, and complement activation. The Fc region of an antibody is contained within the heavy chain constant domain and can interact with Fc receptors located on the cell surface, for example.

[0114] The sequences of antibody genes after construction and somatic mutation are highly diverse, and these diversified genes are 10 It is predicted that these genes encode different types of antibody molecules (Immunoglobulin Genes, 2 nd ed., eds. Jonio et al., Academic Press, San Diego, CA, 1995). Thus, the immune system provides a repertoire of immunoglobulins. The term "repertoire" refers to at least one nucleotide sequence derived in whole or in part from at least one sequence encoding at least one immunoglobulin. The sequence may be generated by in vivo rearrangement of the V, D, and J segments of the heavy chain and the V and J segments of the light chain. Alternatively, the sequence may be generated from a cell in response to, for example, an in vitro stimulus that causes the rearrangement. Alternatively, some or all of the sequence may be obtained by DNA splicing, nucleotide synthesis, mutagenesis, and other methods (see, for example, U.S. Pat. No. 5,565,332). A repertoire may include only one sequence or may include multiple sequences, including those in a genetically diverse collection.

[0115] The term "Fc portion" or "Fc monomer" in the context of the present invention means a polypeptide comprising at least one domain having the function of a CH2 domain and at least one domain having the function of a CH3 domain of an immunoglobulin molecule. As is evident from the term "Fc monomer", a polypeptide comprising these CH domains is a "polypeptide monomer". An Fc monomer may be a polypeptide comprising a fragment of an immunoglobulin constant region excluding at least the first constant region immunoglobulin domain (CH1) of the heavy chain, but maintaining at least a functional part of one CH2 domain and a functional part of one CH3 domain, the CH2 domain being amino-terminal to the CH3 domain. In a preferred embodiment of this definition, an Fc monomer may be a polypeptide constant region comprising a part of an Ig-Fc hinge region, a CH2 region, and a CH3 region, the hinge region being amino-terminal to the CH2 domain. It is envisaged that the hinge region of the present invention promotes dimerization. Such Fc polypeptide molecules can be obtained, for example, but not limited to, by papain digestion of an immunoglobulin region (which of course results in a dimer of two Fc polypeptides). In another aspect of this definition, an Fc monomer can be a polypeptide region that includes a portion of a CH2 region and a CH3 region. Such Fc polypeptide molecules can be obtained, for example, but not limited to, by pepsin digestion of an immunoglobulin molecule. In one embodiment, the polypeptide sequence of an Fc monomer is substantially similar to the Fc polypeptide sequence of an IgG1 Fc region, an IgG2 Fc region, an IgG3 Fc region, an IgG4 Fc region, an IgM Fc region, an IgA Fc region, an IgD Fc region, and an IgE Fc region. (See, for example, Padlan, Molecular Immunology, 31(3), 169-217 (1993)). Because there is some variation between immunoglobulins, and simply for clarity, the Fc monomer refers to the last two heavy chain constant region immunoglobulin domains of IgA, IgD, and IgG, and the last three heavy chain constant region immunoglobulin domains of IgE and IgM. As mentioned, the Fc monomer may also include a flexible hinge N-terminal to these domains. In the case of IgA and IgM, the Fc monomer may include a J chain.In the case of IgG, the Fc portion comprises the immunoglobulin domains CH2 and CH3 and the hinge between the first two domains and CH2. Although the boundaries of the Fc portion may vary, an example of a human IgG heavy chain Fc portion comprising functional hinge, CH2 and CH3 domains can be defined, for example, as comprising residues D231 (of the hinge domain - corresponding to D234 in Table 1 below) to P476, L476 (for IgG4) at the carboxyl terminus of the CH3 domain, respectively, according to the numbering according to Kabat. Two Fc portions or Fc monomers fused to each other via a peptide linker define the third domain of the antigen-binding molecule of the present invention, which can also be defined as an scFc domain.

[0116] In one embodiment of the present invention, it is envisaged that the scFc domains disclosed herein (respectively Fc monomers fused to each other) are comprised only in the third domain of an antigen-binding polypeptide.

[0117] In accordance with the present invention, IgG hinge regions may be identified by similarity with the Kabat numbering set forth in Table 1. In accordance with the above, it is envisaged that for the hinge domain / region of the present invention, the minimum requirement comprises amino acid residues corresponding to the stretch of IgG1 sequence from D231 D234 to P243 according to the Kabat numbering. Similarly, it is envisaged that the hinge domain / region of the present invention comprises or consists of the IgG1 hinge sequence DKTHTCPPCP (SEQ ID NO:) (corresponding to the D234 to P243 section as shown in Table 1 below - variants of said sequence are also envisaged, provided that the hinge region still promotes dimerization). In a preferred embodiment of the present invention, the glycosylation site at Kabat position 314 of the CH2 domain in the third domain of the antigen binding molecule is removed by a N314X substitution (wherein X is any amino acid other than Q). Said substitution is preferably a N314G substitution. In a more preferred embodiment, said CH2 domain further comprises the following substitutions (positions according to Kabat): V321C and R309C (these substitutions introduce intradomain cysteine ​​disulfide bridges at Kabat positions 309 and 321).

[0118] It is also envisaged that the third domain of the antigen-binding molecule of the present invention comprises or consists of, in amino to carboxyl order, DKTHTCPPCP (SEQ ID NO:) (i.e., hinge)-CH2-CH3-linker-DKTHTCPPCP (SEQ ID NO:) (i.e., hinge)-CH2-CH3. The peptide linker of the antigen-binding molecule described above is, in a preferred embodiment, characterized by the amino acid sequence Gly-Gly-Gly-Gly-Ser (i.e., Gly4Ser (SEQ ID NO: 1)), or a polymer thereof (i.e., (Gly4Ser)x (wherein x is an integer equal to or greater than 5 (e.g., 5, 6, 7, 8, etc., or more), with 6 being preferred ((Gly4Ser)6)). The construct may further comprise the substitutions described above: N314X, preferably N314G, and / or further substitutions V321C and R309C. In a preferred embodiment of the antigen binding molecule of the invention as defined hereinbefore, it is envisaged that the second domain binds to an extracellular epitope of the human and / or Macaca CD3 epsilon chain.

[0119] [Table 1]

[0120] In further embodiments of the invention, the hinge domain / region comprises or consists of the IgG2 subtype hinge sequence ERKCCVECPPCP (SEQ ID NO:), the IgG3 subtype hinge sequence ELKTPLDTTHTCPRCP (SEQ ID NO:) or ELKTPLGDTTHTCPRCP (SEQ ID NO:), and / or the IgG4 subtype hinge sequence ESKYGPPCPSCP (SEQ ID NO:). The IgG1 subtype hinge sequence may be the following sequence EPKSCDKTHTCPPCP (as shown in Table 1 and in the SEQ ID NO:). Thus, these core hinge regions are also envisaged in the context of the present invention.

[0121] The locations and sequences of the IgG CH2 and IgG CD3 domains can be identified by similarity using the Kabat numbering set forth in Table 2.

[0122] [Table 2]

[0123] In one embodiment of the invention, the amino acid residues highlighted in bold in the CH3 domain of the first or both Fc monomers are deleted.

[0124] The peptide linker fusing the polypeptide monomers of the third domain ("Fc portion" or "Fc monomer") to one another preferably comprises at least 25 amino acid residues (25, 26, 27, 28, 29, 30, etc.). More preferably, the peptide linker comprises at least 30 amino acid residues (30, 31, 32, 33, 34, 35, etc.). It is also preferred that the linker comprises up to 40 amino acid residues, more preferably up to 35 amino acid residues, most preferably exactly 30 amino acid residues. A preferred embodiment of such a peptide linker is characterized by the amino acid sequence Gly-Gly-Gly-Gly-Ser, i.e., Gly4Ser (SEQ ID NO: 1), or a polymer thereof, i.e., (Gly4Ser)x, where x is an integer equal to or greater than 5 (e.g., 6, 7, or 8). Preferably, the integer is 6 or 7, more preferably, the integer is 6.

[0125] When a linker is used to fuse the first domain with the second domain, or the first domain or the second domain with the third domain, the linker is preferably of sufficient length and sequence to ensure that each of the first and second domains can retain their different binding specificities independently of each other. With respect to the peptide linkers connecting at least two binding domains (or two variable domains) in the antigen-binding molecule of the present invention, those peptide linkers are preferably those that contain only a few amino acid residues (e.g., those that contain 12 amino acid residues or less). Thus, peptide linkers of 12, 11, 10, 9, 8, 7, 6, or 5 amino acid residues are preferred. Conceivable peptide linkers with less than 5 amino acids contain 4, 3, 2, or 1 amino acid, with Gly-rich linkers being preferred. A preferred embodiment of a peptide linker for the fusion of the first and second domains is shown in SEQ ID NO:1. A preferred linker embodiment of the peptide linker for fusing the second and third domains is the (Gly)4-linker, also called a G4-linker.

[0126] A particularly preferred "single" amino acid in connection with one of the above "peptide linkers" is Gly. Thus, said peptide linker may consist of a single amino acid Gly. In a preferred embodiment of the invention, the peptide linker is characterized by the amino acid sequence Gly-Gly-Gly-Gly-Ser (i.e., Gly4Ser (SEQ ID NO: 1)), or a polymer thereof (i.e., (Gly4Ser)x), where x is an integer equal to or greater than 1 (e.g., 2 or 3). Preferred linkers are shown in SEQ ID NOs: 1-12. The characteristics of said peptide linkers, including not promoting secondary structures, are known in the art and are described, for example, in Dall'Acqua et al. (Biochem. (1998) 37, 9266-9273), Cheadle et al. (Mol Immunol (1992) 29, 21-30), and Raag and Whitlow (FASEB (1995) 9(1), 73-80). Moreover, peptide linkers that do not promote any secondary structure are preferred. The mutual linkage of the domains can be provided, for example, by genetic engineering as described in the Examples. Methods for preparing fused and operably linked bispecific single chain constructs and expressing them in mammalian cells or bacteria are known in the art (e.g., WO 99 / 54440 or Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 2001).

[0127] In a preferred embodiment of the antigen-binding molecule of the present invention, the first domain and the second domain form an antigen-binding molecule of a type selected from the group consisting of (scFv)2, scFv-single domain mAb, diabody, and oligomers of any of these types.

[0128] According to a particularly preferred embodiment, and as described in the accompanying examples, the first and second domains of the antigen-binding molecule of the present invention are "bispecific single-chain antigen-binding molecules", more preferably bispecific "single-chain Fvs" (scFvs). Although the two domains of the Fv fragment, VL and VH, are encoded by separate genes, they can be linked by a synthetic linker as described herein above, which allows them to be produced using recombinant methods as a single protein chain in which the VL and VH regions are paired to form a monovalent molecule; see, for example, Huston et al. (1988) Proc. Natl. Acad. Sci USA 85:5879-5883). These antibody fragments are obtained using conventional techniques known to those skilled in the art, and the fragments are evaluated for function in the same manner as complete or full-length antibodies. Thus, a single chain variable fragment (scFv) is a fusion protein of the variable domains of the heavy (VH) and light (VL) chains of immunoglobulins, usually linked by a short linker peptide as described herein. The linker is usually rich in glycine for flexibility and serine or threonine for solubility, and can either link the N-terminus of VH to the C-terminus of VL or vice versa. The protein retains the specificity of the original immunoglobulin despite the removal of the constant regions and the introduction of the linker.

[0129] Bispecific single-chain antigen-binding molecules are known in the art and have been described in WO 99 / 54440; Mack, J. Immunol. (1997), 158, 3965-3970; Mack, PNAS, (1995), 92, 7021-7025; Kufer, Cancer Immunol. Immunother., (1997), 45, 193-197; Loeffler, Blood, (2000), 95, 6, 2098-2103; Bruehl, Immunol., (2001), 166, 2420-2426; Kipriyanov, J. Mol. Biol., (1999), 293, 41-56. Techniques described for the production of single chain antibodies (see, inter alia, U.S. Pat. No. 4,946,778; Kontermann and Duebel (2010), supra; and Little (2009), supra) can be adapted to generate single chain antigen-binding molecules that specifically recognize a target of choice.

[0130] Bivalent (also called divalent) or bispecific single chain variable fragments (bi-scFv or di-scFv with the format (scFv)2) can be created by linking two scFv molecules (e.g., using a linker as described previously herein). If these two scFv molecules have the same binding specificity, the resulting (scFv)2 molecule is preferably called bivalent (i.e., it has a valency of two for the same target epitope). If these two scFv molecules have different binding specificities, the resulting (scFv)2 molecule is preferably called bispecific. Linking can be done by creating a single peptide chain with two VH and two VL regions to generate tandem scFvs (see, for example, Kufer P. et al., (2004) Trends in Biotechnology 22(5):238-244). Another possibility is to create scFv molecules with a linker peptide that is too short (e.g., about 5 amino acids) to allow the two variable regions to fold together, allowing the scFvs to dimerize, a form known as a diabody (see, e.g., Hollinger, Philipp et al., (July 1993) Proceedings of the National Academy of Sciences of the United States of America 90(14):6444-8).

[0131] Consistent with the present invention, either the first domain, the second domain, or the first and second domains may comprise a single domain antibody, a variable domain of a single domain antibody, or at least the CDRs of a single domain antibody, respectively. Single domain antibodies contain only one (monomeric) antibody variable domain that can selectively bind to a specific antigen independent of other V regions or domains. The first single domain antibodies were generated from heavy chain antibodies found in camels, which consist of V H The cartilaginous fish also have heavy chain antibodies (IgNAR) from which the V NARSingle domain antibodies, called fragments, can be obtained. An alternative approach is to split the dimeric variable domain from a common immunoglobulin, for example from human or rodent origin, into monomers, thereby obtaining VH or VL as single domain Abs. Most research on single domain antibodies is currently based on heavy chain variable domains, but nanobodies derived from light chains have also been shown to specifically bind target epitopes. Examples of single domain antibodies are called sdAbs, nanobodies, or single variable domain antibodies.

[0132] Therefore, (single domain mAb)2 is V H , V L , V H H and V NAR A monoclonal antigen-binding molecule is a monoclonal antigen-binding molecule composed of (at least) two single domain monoclonal antibodies individually selected from the group comprising: a) a single domain antibody as defined above, b) a single domain antibody as defined above, c) a single domain antibody as defined above, and d) a single domain antibody as defined above, e). The linker is preferably in the form of a peptide linker. Similarly, an "scFv-single domain mAb" is a monoclonal antigen-binding molecule composed of at least one single domain antibody as defined above and one scFv molecule as defined above. Again, the linker is preferably in the form of a peptide linker.

[0133] Whether an antigen-binding molecule binds to another given antigen-binding molecule in competition can be measured by a competitive assay, such as a competitive ELISA or a cell-based competitive assay. Avidin-conjugated microparticles (beads) can also be used. Each of these beads can be used as a substrate when reacting with biotinylated proteins, as with an avidin-coated ELISA plate, on which the assay can be carried out. Antigen is coated on the beads, and then pre-coated with the first antibody. A secondary antibody is added to confirm any further binding. Possible means for reading include flow cytometry.

[0134] T cells or T lymphocytes are a type of lymphocyte (itself a type of white blood cell) that plays a central role in cell-mediated immunity. There are several subsets of T cells, each with different functions. T cells can be distinguished from other lymphocytes, such as B cells and NK cells, by the presence of T cell receptors (TCRs) on their cell surface. The TCR is responsible for the recognition of antigens bound to major histocompatibility complex (MHC) molecules and is composed of two different protein chains. In 95% of T cells, the TCR consists of an alpha (α) chain and a beta (β) chain. When the TCR binds to an antigen peptide and MHC (peptide / MHC complex), the T lymphocyte is activated through a series of biochemical events mediated by associated enzymes, co-receptors, specialized adaptor molecules, and activated or released transcription factors.

[0135] The CD3 receptor complex is a protein complex and is composed of four chains. In mammals, this complex contains the CD3γ (gamma) chain, the CD3δ (delta) chain, and two CD3ε (epsilon) chains. These chains associate with the T cell receptor (TCR) and the so-called ζ (zeta) chain to form the T cell receptor CD3 complex, which generates activation signals in T lymphocytes. The CD3γ (gamma), CD3δ (delta), and CD3ε (epsilon) chains are very closely related cell surface proteins of the immunoglobulin superfamily that contain a single extracellular immunoglobulin domain. The intracellular tail of the CD3 molecule contains a single conserved motif known as the immunoreceptor tyrosine-based activation motif or ITAM for short, which is essential for the signaling ability of the TCR. The CD3 epsilon molecule is a polypeptide encoded by the CD3E gene, which in humans is present on chromosome 11. The most preferred epitope of CD3 epsilon is comprised within the range of amino acid residues 1 to 27 of the extracellular domain of human CD3 epsilon. It is typically and advantageously expected that the antigen-binding molecule according to the present invention will not show much undesirable non-specific T cell activation in certain immunotherapies, which translates into a low risk of side effects.

[0136] Lysis of redirected target cells via recruitment of T cells by multispecific (at least bispecific) antigen-binding molecules involves the formation of a cytolytic synapse and delivery of perforin and granzymes. Bound T cells are capable of continuous target cell lysis and are not subject to immune evasion mechanisms that prevent peptide antigen processing and presentation or clonal T cell differentiation (see, e.g., WO 2007 / 042261).

[0137] The cytotoxicity mediated by the antigen-binding molecules of the present invention can be measured in various ways. Effector cells can be, for example, stimulated enriched (human) CD8 positive T cells or unstimulated (human) peripheral blood mononuclear cells (PBMCs). If the target cells are of macaque origin, or express or are transfected with macaque CD20 or CD22 to which the first domain binds, the effector cells should also be of macaque origin, such as a macaque T cell line (e.g., 4119LnPx). The target cells should express (at least the extracellular domain of) CD20 or CD22 (e.g., human or macaque CD20 or CD22). The target cells can be a cell line (e.g., CHO) that is stably or transiently transfected with CD20 or CD22 (e.g., human or macaque CD20 or CD22). Usually, EC 50 Values ​​are expected to be lower for target cell lines expressing high levels of CD20 or CD22 on the cell surface. Effector to target cell (E:T) ratios are usually around 10:1 but may vary. 51 Cytotoxicity may be measured by Cr release assay (incubation time of about 18 hours) or FACS-based cytotoxicity assay (incubation time of about 48 hours). Modification of assay incubation time (cytotoxic response) is also possible. Other methods for measuring cytotoxicity are known to those skilled in the art and include MTT or MTS assays, ATP-based assays including bioluminescence assays, sulforhodamine B (SRB) assays, WST assays, clonogenic assays, and ECIS techniques.

[0138] The cytotoxic activity mediated by the CD20 and CD22xCD3 bispecific antigen-binding molecules of the present invention is preferably measured in a cell-based cytotoxicity assay. 51 It can also be measured by Cr release assay. 50 The EC50 value corresponds to the half maximal effective concentration (the concentration of an antigen-binding molecule that induces a cytotoxic response halfway between the baseline and maximum). Preferably, the EC50 value corresponds to the half maximal effective concentration (the concentration of an antigen-binding molecule that induces a cytotoxic response halfway between the baseline and maximum). 50 The value is ≦5000 pM or ≦4000 pM, more preferably ≦3000 pM or ≦2000 pM, even more preferably ≦1000 pM or ≦500 pM, even more preferably ≦400 pM or ≦300 pM, even more preferably ≦200 pM, even more preferably ≦100 pM, even more preferably ≦50 pM, even more preferably ≦20 pM or ≦10 pM, and most preferably ≦5 pM.

[0139] In various assays, the EC 50 Values ​​can be measured. Stimulated / enriched CD8 + When T cells were used as effector cells, EC 50 Those skilled in the art will recognize that values ​​may be expected to be lower. 50 Values ​​would be expected to be lower if the target cells express high amounts of CD20 and CD22 compared to rats with low target expression. For example, stimulated / enriched human CD8 + When T cells are used as effector cells (and either CD20 and CD22 transfected cells, such as CHO cells, or CD20 and CD22 positive human cell lines are used as target cells), the EC of CD20 or CD22 bispecific antigen-binding molecules is 50The value is preferably ≦1000 pM, more preferably ≦500 pM, even more preferably ≦250 pM, even more preferably ≦100 pM, even more preferably ≦50 pM, even more preferably ≦10 pM, and most preferably ≦5 pM. When human PBMCs are used as effector cells, the EC 50 The value is preferably ≦5000 pM or ≦4000 pM (especially when the target cells are CD20 or CD22 positive human cell lines), more preferably ≦2000 pM, more preferably ≦1000 pM or ≦500 pM, even more preferably ≦200 pM, even more preferably ≦150 pM, even more preferably ≦100 pM, and most preferably ≦50 pM. When a macaque T cell line such as LnPx4119 is used as the effector cell and a macaque CD20 or CD22 transfected cell line such as CHO cells is used as the target cell line, the EC 50 The value is preferably ≦2000 pM or ≦1500 pM, more preferably ≦1000 pM or ≦500 pM, even more preferably ≦300 pM or ≦250 pM, even more preferably ≦100 pM, and most preferably ≦50 pM.

[0140] Preferably, the CD20 and CD22×CD3 bispecific antigen binding molecules of the present invention do not induce / mediate or essentially do not induce / mediate lysis of CD20 and CD22 negative cells, such as CHO cells. The terms "do not induce lysis", "essentially do not induce lysis", "do not mediate lysis" or "essentially do not mediate lysis" mean that the antigen binding molecules of the present invention do not induce or mediate lysis of more than 30%, preferably more than 20%, more preferably more than 10%, particularly preferably more than 9%, 8%, 7%, 6% or 5% of CD20 or CD22 negative cells, when the lysis of CD020 or CD22 positive human cell lines is taken as 100%. This usually applies to concentrations of antigen binding molecules up to 500 nM. The skilled person knows how to measure cell lysis without further effort. Furthermore, specific instructions for measuring cell lysis are taught herein.

[0141] The difference in cytotoxic activity between the monomeric and dimeric isoforms of the individual CD20 and CD22×CD3 bispecific antigen-binding molecules is referred to as the "potency gap." This potency gap can be measured, for example, by comparing the EC 50 Values ​​and EC of dimeric form 50 The potency gap of the CD20 and CD22xCD3 bispecific antigen binding molecules of the invention is preferably ≦5, more preferably ≦4, even more preferably ≦3, even more preferably ≦2, and most preferably ≦1.

[0142] The first and / or second (or any further) binding domains of the antigen binding molecules of the present invention are preferably cross-species specific for members of the mammalian order of primates. Cross-species specific CD3 binding domains are described, for example, in WO 2008 / 119567. According to one embodiment, the first and / or second binding domain, in addition to binding to human CD20 and CD22 and human CD3, will also bind to CD20 and CD22 / CD3, respectively, of primates, including, but not limited to, New World primates (e.g., Callithrix jacchus, Saguinus Oedipus, or Saimiri sciureus), Old World primates (e.g., baboons and macaques), gibbons, and non-human hominins.

[0143] In one embodiment of the antigen-binding molecule of the present invention, the first domain binds to human CD20 and CD22 and further binds to macaque CD20 and CD22, such as cynomolgus monkey (Macaca fascicularis) CD20 and CD22, more preferably to macaque CD20 and CD22 expressed on the surface of cells, such as CHO or 293 cells. The affinity of the first domain for CD20 and CD22 (preferably human CD20 and CD22) is preferably ≦100 nM or ≦50 nM, more preferably ≦25 nM or ≦20 nM, more preferably ≦15 nM or ≦10 nM, even more preferably ≦5 nM, even more preferably ≦2.5 M or ≦2 M, even more preferably ≦1 nM, even more preferably ≦0.6 nM, even more preferably ≦0.5 nM, and most preferably ≦0.4 nM. The affinity can be measured, for example, by BIAcore assay or Scatchard assay. Other methods of determining affinity are known to those skilled in the art. The affinity of the first domain for macaque CD20 and CD22 is preferably ≦15 nM, more preferably ≦10 nM, even more preferably ≦5 nM, even more preferably ≦1 nM, even more preferably ≦0.5 nM, even more preferably ≦0.1 nM, and most preferably ≦0.05 nM or even ≦0.01 nM.

[0144] Preferably, the affinity gap of the antigen-binding molecule of the present invention for the binding of macaque CD20 and CD22 to human CD20 and CD22 [ma CD20 and CD22:hu CD20 and CD22] (determined, for example, by BiaCore or Scatchard analysis) is <100, preferably <20, more preferably <15, even more preferably <10, even more preferably <8, more preferably <6, and most preferably <2. The preferred range of the affinity gap of the antigen-binding molecule of the present invention for the binding of macaque CD20 and CD22 to human CD20 and CD22 is 0.1 to 20, more preferably 0.2 to 10, even more preferably 0.3 to 6, even more preferably 0.5 to 3 or 0.5 to 2.5, and most preferably 0.5 to 2 or 0.6 to 2.

[0145] The third binding domain of the antigen binding molecule of the present invention binds to human CD3 epsilon and / or Macaca CD3 epsilon. In a preferred embodiment, the second domain further binds to CD3 epsilon of Callithrix jacchus, Saguinus Oedipus, or Saimiri sciureus. Both Callithrix jacchus and Saguinus oedipus are New World primates belonging to the Callitrichidae family, while Saimiri sciureus is a New World primate belonging to the Cebidae family. The binding domain may preferably be referred to as "I2C" or "I2C0" in Table 5.

[0146] With respect to the antigen binding molecule of the present invention, it is preferred that the third binding domain binding to an extracellular epitope of the human and / or Macaca CD3 epsilon chain comprises a VL region comprising CDR-L1, CDR-L2, and CDR-L3 selected from: (a) SEQ ID NOs: 392 to 394; and (b) SEQ ID NOs: 395 to 397.

[0147] In a further preferred embodiment of the antigen binding molecule of the present invention, the third domain binding to an extracellular epitope of the human and / or Macaca CD3 epsilon chain comprises a VH region comprising CDR-H1, CDR-H2, and CDR-H3 selected from: (a) SEQ ID NOs: 400 to 402; and (b) SEQ ID NOs: 403 to 405.

[0148] In a preferred embodiment of the antigen-binding molecule of the present invention, the above three groups of VL CDRs are combined with the above ten groups of VH CDRs in the third binding domain to form groups comprising CDR-L1 to 3 and CDR-H1 to 3, respectively.

[0149] In the antigen-binding molecule of the present invention, the third domain that binds to CD3 preferably comprises a VL region selected from the group consisting of those shown in SEQ ID NOs: 17, 21, 35, 39, 53, 57, 71, 75, 89, 93, 107, 111, 125, 129, 143, 147, 161, 165, 179, or 183 of WO 2008 / 119567, or those shown in SEQ ID NO: 13 of the present invention.

[0150] The third domain that binds to CD3 comprises a VH region selected from the group consisting of those set forth in SEQ ID NO: 15, 19, 33, 37, 51, 55, 69, 73, 87, 91, 105, 109, 123, 127, 141, 145, 159, 163, 177, or 181 of WO 2008 / 119567 or set forth in SEQ ID NO: 14.

[0151] More preferably, the antigen binding molecule of the present invention is characterized by a CD3-binding third domain comprising a VL region and a VH region selected from the group consisting of: (a) a VL region as set forth in SEQ ID NO: 17 or 21 of WO 2008 / 119567, and a VH region as set forth in SEQ ID NO: 15 or 19 of WO 2008 / 119567; (b) a VL region as set forth in SEQ ID NO: 35 or 39 of WO 2008 / 119567, and a VH region as set forth in SEQ ID NO: 33 or 37 of WO 2008 / 119567; (c) the VL region as set forth in SEQ ID NO: 53 or 57 of WO 2008 / 119567, and the VH region as set forth in SEQ ID NO: 51 or 55 of WO 2008 / 119567; (d) a VL region as set forth in SEQ ID NO: 71 or 75 of WO 2008 / 119567, and a VH region as set forth in SEQ ID NO: 69 or 73 of WO 2008 / 119567; (e) a VL region as set forth in SEQ ID NO: 89 or 93 of WO 2008 / 119567, and a VH region as set forth in SEQ ID NO: 87 or 91 of WO 2008 / 119567; (f) the VL region as set forth in SEQ ID NO: 107 or 111 of WO 2008 / 119567, and the VH region as set forth in SEQ ID NO: 105 or 109 of WO 2008 / 119567; (g) the VL region as set forth in SEQ ID NO: 125 or 129 of WO 2008 / 119567, and the VH region as set forth in SEQ ID NO: 123 or 127 of WO 2008 / 119567; (h) the VL region as set forth in SEQ ID NO: 143 or 147 of WO 2008 / 119567, and the VH region as set forth in SEQ ID NO: 141 or 145 of WO 2008 / 119567; (i) the VL region as set forth in SEQ ID NO: 161 or 165 of WO 2008 / 119567, and the VH region as set forth in SEQ ID NO: 159 or 163 of WO 2008 / 119567; and (j) a VL region as set forth in SEQ ID NO: 179 or 183 of WO 2008 / 119567, and a VH region as set forth in SEQ ID NO: 177 or 181 of WO 2008 / 119567.

[0152] A third domain that binds to CD3, comprising the VL region shown in SEQ ID NO: 13 and the VH region shown in SEQ ID NO: 14, is also preferred in the context of the antigen-binding molecule of the present invention.

[0153] According to a preferred embodiment of the antigen-binding molecule of the present invention, the first domain and / or the third domain have the following format: the pair of VH and VL domains is in the format of a single chain antibody (scFv). The VH and VL domains are arranged in the order of VH-VL or VL-VH. It is preferred that the VH domain is arranged at the N-terminus of the linker sequence, and the VL domain is arranged at the C-terminus of the linker sequence.

[0154] A preferred embodiment of the above-mentioned antigen-binding molecule of the present invention is characterized by a CD3-binding third domain comprising an amino acid sequence selected from the group consisting of SEQ ID NOs: 23, 25, 41, 43, 59, 61, 77, 79, 95, 97, 113, 115, 131, 133, 149, 151, 167, 169, 185, or 187 of WO2008 / 119567, or as set forth in SEQ ID NO: 15.

[0155] The present invention relates to nucleic acids having SEQ ID NOs: 673, 676, 679, 682, 685, 688, 691, 694, 697, 700, 703, 706, 709, 712, 715, 718, 721, 724, 727, 730, 733, 736, 739, 742, 745, 748, 751, 754, 757, 760, 763, 766, 769, 772, 775, 778, 781, 784, 787, 790, 793, 796, 799, 802, 805, 808, 811, 814, 817, 820, 823, 826, 829, 832, 835, 838, 841, 844, 847, 850, 853, 856, 859, 862, 865, 868, 871, 1437, 1440, 1443, 1446, 1449, 1452, 1455, 1458, 1461, 1464, 1467, 1470, 1473, 1476, 1479, 1482, 1485, 1488, 1499, 1 667, 1670, 1673, 1676, 1679, 1682, 1685, 1688, 1691, 1694, 1697, 1700, 1703, 1706, 1709, 1712, 1715, 1718, 1721, 1724, 1727, 1730, 1733, 1736, 1739, 1 742, 1745, 1748, 1751, 1754, 1757, 1760, 1763, 1766, 1769, 1772, 1775, 1778, 1781, 1784, 1787, 1790, 1793, 1796, 1799, 1802, 1805, 1808, 1811, 1814, 1817, 1820, 1823, 1826, 1829, 1838, 1851, 1864, 1877, 1890, 1903, 1916, 1933, 1946, 1959, 1972, 1985, 1998, 2011, 2024, 2037, 2050, 2063, 2076, 2089, 2102, 2115, 2128, 2141, 2154, 2167, 2180, 2194, 2206, 2219, 2232, 2245, 2258, 2262, 2270, 2271, 2280, 2281, 2290, 2291, 2300, 2301, 2310, 2311, 2320, 2321, 2330, 2331, 2340, 2341, 2350, 2351, 2360, 2361, 2370, 2371, 2380, 2381, 2390, 2391, 2400, 2401, 2410, 2411, 2420, 2421, 2430, 2431, 2440, 2441,2450、2451、2460、2461、2470、2471、2480、2481、2490、2491、2500、2501、2510、2511、2520、2521、2530、2531、2540、2541、2550、2551、2560、2561、2570、2571、2580、2581、2590、2591、2600、2601、2610、2611、2620、2621、2630、2631、2640、2641、2650、2651、2660、2661、2670、2671、2680、2681、2690、2691、2700、2701、2710、2711、2720、2721、2730、2731、2740、2741、2750、2751、2760、2761、2770、2771、2780、2781、2790、2791、2800、2801、2810、2811、2820、2821、2830、2831、2840、2841、2850、2851、2860、2861、2870、2871、2880、2881、2890、2891、2900、2901、2910、2911、2920、2921、2930、2931、2940、2941、2950、2951、2960、2961、2970、2971、2980、2981、2990、2991、3000、3001、3010、3011、3020、3021、3030、3031、3040、3041、3050、3051、3060、3061、3070、3071、3080、3081、3090、3091、3100、3101、3110、3111、3120、3121、3130、3131、3140、3141、3150、3151、3160、3161、3170、3171、3180、3181、3190、3191、3200、3201、3210、3211、3220、3221、3231、3240、3241、3250、3251、3260、3261、3270、3271、3280、3281、3290、3291、3300、3301、3310、3311、3320、3321、3330、3331、3340、3341、3344、3345、3356、3367、3378、3389、3400、3411、3422、3433、3444、3455、3466、3477、3488、3499、3510、3521、3532、3543、3554、3565, 3576, 3579, 3582, 3585, 3588, 3591, 3594, 3597, 3600, 3603, 3606, 3609, 3612, 3615, 3618, 3621, 3624, 3627, 3630, 3633, 3636, 3639, 3642, 3645, 3648, 3651, 3654, 3657, 3660, 3663, 3666, 3669, 3672, 3675, 3678, 3689, 3700, 3704, 3705, 3708, 3709, 3710, 3 The present invention further provides an antigen-binding molecule (fully bispecific antigen-binding molecule) that comprises or has an amino acid sequence selected from the group consisting of any of the following sequences: 3711, 3722, 3733, 3736, 3739, 3744, 3747, 3748, 3756, 3757, 3761, and 3762, preferably 1437, or an antigen-binding molecule that has an amino acid sequence that is at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% identical to said sequences.

[0156] Covalent modifications of antigen-binding molecules are also included within the scope of the present invention, which are generally, but not necessarily, performed post-translationally. For example, some types of covalent modifications of antigen-binding molecules are introduced into the molecule by reacting specific amino acid residues of the antigen-binding molecule with organic derivatizing agents capable of reacting with selected side chains or N- or C-terminal residues.

[0157] Cysteinyl residues most commonly are reacted with α-haloacetates (and corresponding amines), such as chloroacetic acid or chloroacetamide to give carboxymethyl or carboxyamidomethyl derivatives. Cysteinyl residues are also derivatized by reaction with bromotrifluoroacetone, α-bromo-β-(5-imidozoyl)propionic acid, chloroacetyl phosphate, N-alkylmaleimides, 3-nitro-2-pyridyl disulfide, methyl 2-pyridyl disulfide, p-chloromercuribenzoic acid, 2-chloromercuri-4-nitrophenol, or chloro-7-nitrobenzo-2-oxa-1,3-diazole.

[0158] Histidyl residues are derivatized by reaction with diethylpyrocarbonate at pH 5.5-7.0 because this agent is relatively specific for the histidyl side chain. Para-bromophenacyl bromide is also useful, and the reaction is preferably carried out in 0.1 M sodium cacodylate at pH 6.0. Lysinyl and amino terminal residues react with succinic or other carboxylic acid anhydrides. Derivatization with these agents has the effect of reversing the charge of lysinyl residues. Other suitable reagents for derivatizing alpha-amino-containing residues include imidoesters such as methyl picolinimidate; pyridoxal phosphate; pyridoxal; chloroborohydrides; trinitrobenzenesulfonic acid; O-methylisourea; 2,4-pentanedione; and transaminase-catalyzed reactions with glyoxylate.

[0159] Arginyl residues are modified by reaction with one or several conventional reagents, among them phenylglyoxal, 2,3-butanedione, 1,2-cyclohexanedione, and ninhydrin. Due to the high pKa of the guanidine functional group, derivatization of arginine residues requires that the reaction be carried out under alkaline conditions. Moreover, these reagents can react with lysine groups and the arginine epsilon-amino group.

[0160] The specific modification of tyrosyl residues may be undertaken for the introduction of spectral labels into tyrosyl residues by reaction with aromatic diazonium compounds or tetranitromethane. Most commonly, N-acetylimidazole and tetranitromethane are used to form O-acetyl tyrosyl species and 3-nitro derivatives, respectively. 125 I or 131 The chloramine T method described above, in which tyrosyl residues are iodinated with I to prepare labeled proteins for use in radioimmunoassay, is preferred.

[0161] Carboxyl side groups (aspartyl or glutamyl) are selectively modified by reaction with carbodiimides (R'-N=C=N-R'), where R and R' are optionally different alkyl groups, such as 1-cyclohexyl-3-(2-morpholinyl-4-ethyl)carbodiimide or 1-ethyl-3-(4-azonia-4,4-dimethylpentyl)carbodiimide. Furthermore, aspartyl and glutamyl residues are converted to asparaginyl and glutaminyl residues by reaction with ammonium ions.

[0162] Derivatization with bifunctional agents is useful for crosslinking the antigen-binding molecules of the present invention to water-insoluble support matrices or surfaces for use in various methods. Commonly used crosslinking agents include, for example, 1,1-bis(diazoacetyl)-2-phenylethane, glutaraldehyde, N-hydroxysuccinimide esters, such as esters with 4-azidosalicylic acid, homobifunctional imidoesters including disuccinimidyl esters, such as 3,3-dithiobis(succinimidyl propionate), and bifunctional maleimides, such as bis-N-maleimido-1,8-octane. Derivatizing agents such as methyl-3-[(p-azidophenyl)dithio]propioimidate provide photoactivatable intermediates capable of forming crosslinks in the presence of light. Alternatively, reactive water-insoluble matrices such as cyanogen bromide-activated carbohydrates and reactive substrates described in U.S. Pat. Nos. 3,969,287; 3,691,016; 4,195,128; 4,247,642; 4,229,537; and 4,330,440 are utilized for protein immobilization.

[0163] Glutaminyl and asparaginyl residues are frequently deamidated to the corresponding glutamyl and aspartyl residues, respectively, or they are deamidated under mildly acidic conditions. Either form of these residues is included within the scope of the invention.

[0164] Other modifications include hydroxylation of proline and lysine, phosphorylation of the hydroxyl group of seryl or threonyl residues, methylation of the α-amino groups of lysine, arginine, and histidine side chains (TECreighton, Proteins: Structure and Molecular Properties, W.H. Freeman & Co., San Francisco, 1983, pp. 79-86), acetylation of the N-terminal amine, and amidation of any C-terminal carboxyl group.

[0165] Another type of covalent modification of antigen-binding molecules included within the scope of the present invention includes altering the glycosylation pattern of a protein. As is known in the art, the glycosylation pattern can depend on both the sequence of the protein (e.g., the presence or absence of specific glycosylated amino acid residues, discussed below) or the host cell or organism in which the protein is produced. Specific expression systems are discussed below.

[0166] Glycosylation of polypeptides is usually either N-linked or O-linked. N-linked refers to the attachment of a sugar chain to the side chain of an asparagine residue. The tripeptide sequences asparagine-X-serine and asparagine-X-threonine, where X is any amino acid except proline, are the recognition sequences for enzymatic attachment of a sugar chain to the asparagine side chain. Thus, the presence of either of these tripeptide sequences in a polypeptide creates a potential glycosylation site. O-linked glycosylation refers to the addition of a single sugar, N-acetylgalactosamine, galactose, or xylose, to a hydroxyamino acid, most commonly serine or threonine, although 5-hydroxyproline or 5-hydroxylysine may also be used.

[0167] Addition of glycosylation sites to an antigen-binding molecule can be conveniently accomplished by modifying the amino acid sequence to contain one or more of the above tripeptide sequences (for N-linked glycosylation sites). Modification can also be made by the addition to or substitution by one or more serine or threonine residues from the starting sequence (for O-linked glycosylation sites). Briefly, it is preferred to modify the amino acid sequence of an antigen-binding molecule by modification at the DNA level, in particular by mutating the DNA encoding the polypeptide at preselected bases to generate codons that will be translated into the desired amino acid.

[0168] Another means of increasing the number of carbohydrate chains on an antigen-binding molecule is by chemical or enzymatic coupling of glycosides to the protein. These procedures are advantageous in that they do not require production of the protein in a host cell that has glycosylation capabilities for N-linked and O-linked glycosylation. Depending on the mode of attachment used, sugars can be added to (a) arginine and histidine, (b) free carboxyl groups, (c) free sulfhydryl groups such as those of cysteine, (d) free hydroxyl groups such as those of serine, threonine, or hydroxyproline, (e) aromatic residues such as those of phenylalanine, tyrosine, or tryptophan, or (f) the amide group of glutamine. These methods are described in WO 87 / 05330 and Aplin and Wriston, 1981, CRC Crit. Rev. Biochem., pp. 259-306.

[0169] Removal of carbohydrate moieties present on the starting antigen-binding molecule can be accomplished chemically or enzymatically. Chemical deglycosylation requires exposure of the protein to the compound trifluoromethanesulfonic acid or an equivalent compound. This treatment cleaves most or all sugars except the linking sugar (N-acetylglucosamine or N-acetylgalactosamine) while leaving the polypeptide intact. Chemical deglycosylation is described in Hakimuddin et al., 1987, Arch. Biochem. Biophys. 259:52 and Edge et al., 1981, Anal. Biochem. 118:131. Enzymatic cleavage of carbohydrate moieties on polypeptides can be achieved by the use of various endo- and exoglycosidases as described in Thotakura et al., 1987, Meth. Enzymol. 138:350. Glycosylation at potential glycosylation sites can be prevented by the use of the compound tunicamycin described in Duskin et al., 1982, J. Biol. Chem. 257: 3105. Tunicamycin blocks the formation of protein-N-glycosidic bonds.

[0170] Other modifications of antigen-binding molecules are also contemplated herein.For example, another type of covalent modification of antigen-binding molecules includes linking antigen-binding molecules to various non-proteinaceous polymers, including but not limited to various polyols, such as polyethylene glycol, polypropylene glycol, polyoxyalkylenes, or copolymers of polyethylene glycol and polypropylene glycol, in the manner described in U.S. Patent No. 4,640,835; No. 4,496,689; No. 4,301,144; No. 4,670,417; No. 4,791,192, or No. 4,179,337.In addition, as known in the art, amino acid substitutions can be made at various positions in antigen-binding molecules to facilitate the addition of polymers such as PEG.

[0171] In some embodiments, the covalent modification of the antigen-binding molecule of the present invention comprises the addition of one or more labels. The labeling group can be conjugated to the antigen-binding molecule via a spacer arm of various lengths to reduce potential steric hindrance. Various methods for labeling proteins are known in the art and can be used in carrying out the present invention. The term "label" or "labeling group" refers to any detectable label. Generally, labels are divided into various classes depending on the assay in which they will be detected, including but not limited to the following examples: a) Radioisotopes or radionuclides (e.g. 3 H, 14 C. 15 N, 35 S, 89 Zr, 90 Y, 99 Tc, 111 In, 125 I, 131 Isotopic labels, which may be radioactive or heavy isotopes such as I) b) Magnetic labels (e.g., magnetic particles) c) redox-active moieties d) optical dyes (e.g., but not limited to, chromophores, fluorophores, and fluorophores), such as fluorescent groups (e.g., FITC, rhodamine, lanthanide fluorophores), chemiluminescent groups, and fluorophores, which can be either "small molecule" fluorophores or proteinaceous fluorophores; e) Enzyme groups (e.g., horseradish peroxidase, β-galactosidase, luciferase, alkaline phosphatase) f) Biotinylation Group g) A predetermined polypeptide epitope recognized by a secondary reporter (eg, a leucine zipper pair sequence, a binding site for a secondary antibody, a metal binding domain, an epitope tag, etc.).

[0172] "Fluorescent label" refers to any molecule that can be detected by its inherent fluorescent properties. Suitable fluorescent labels include, but are not limited to, fluorescein, rhodamine, tetramethylrhodamine, eosin, erythrosine, coumarin, methylcoumarins, pyrene, malachite green, stilbene, Lucifer Yellow, Cascade BlueJ, Texas Red, IAEDANS, EDANS, BODIPY FL, LC Red 640, Cy 5, Cy 5.5, LC Red 705, Oregon Green, Alexa-Fluor dyes (Alexa Fluor 350, Alexa Fluor 430, Alexa Fluor 488, Alexa Fluor 546, Alexa Fluor 568, Alexa Fluor 594, Alexa Fluor 633, Alexa Fluor 660, Alexa Fluor 680), Cascade Blue, Cascade Yellow, and R-Phycoerythrin (PE) (Molecular Probes, Eugene, OR), FITC, rhodamine and Texas Red (Pierce, Rockford, IL), Cy5, Cy5.5, Cy7 (Amersham Life Science, Pittsburgh, PA). Suitable optical dyes, including fluorophores, are described in Molecular Probes Handbook by Richard P. Haugland.

[0173] Suitable proteinaceous fluorescent labels include, but are not limited to, GFPs of Renilla, Ptilosarcus, or Aequorea species (Chalfie et al., 1994, Science 263:802-805), green fluorescent proteins including EGFP (Clontech Laboratories, Inc., Genbank Accession No. U55762), blue fluorescent proteins (BFP, Quantum Biotechnologies, Inc. 1801 de Maisonneuve Blvd. West, 8th Floor, Montreal, Quebec, Canada H3H 1J9; Stauber, 1998, Biotechniques 24:462-471; Heim et al., 1996, Curr. Biol. 6:178-182), enhanced yellow fluorescent protein (EYFP, Clontech Laboratories, Inc.), luciferase (Ichiki et al., 1993, J. Immunol. 150:5408-5417), β-galactosidase (Nolan et al. al., 1988, Proc. Natl. Acad. Sci. USA 85:2603-2607), and Renilla (WO 92 / 15673; WO 95 / 07463; WO 98 / 14605; WO 98 / 26277; WO 99 / 49019; U.S. Pat. Nos. 5,292,658; 5,418,155; 5,683,888; 5,741,668; 5,777,079; 5,804,387; 5,874,304; 5,876,995; 5,925,558).

[0174] The antigen-binding molecules of the present invention may also include additional domains, for example, useful for the isolation of the molecule or for the tailored pharmacokinetic profile of the molecule. The domains useful for the isolation of the antigen-binding molecule may be selected from peptide motifs or secondarily introduced moieties that can be captured in an isolation method (e.g., an isolation column). Non-limiting examples of such additional domains include peptide motifs known as Myc tags, HAT tags, HA tags, TAP tags, GST tags, chitin-binding domains (CBD tags), maltose-binding protein (MBP tags), Flag tags, Strep tags, and variants thereof (e.g., StrepII tags), and His tags. All of the antigen-binding molecules disclosed herein may include a His tag domain, commonly known as a repeat of consecutive His residues (preferably five, more preferably six His residues (hexahistidine)) in the amino acid sequence of the molecule. The His tag may be, for example, located at the N-terminus or C-terminus of the antigen-binding molecule, and is preferably located at the C-terminus. Most preferably, a hexa-histidine tag (HHHHHH) (SEQ ID NO: 16) is linked to the C-terminus of the antigen-binding molecule of the present invention via a peptide bond. In addition, the PLGA-PEG-PLGA conjugate system may be combined with a polyhistidine tag for sustained release applications and improved pharmacokinetic profile.

[0175] Amino acid sequence modifications of the antigen-binding molecules described herein are also contemplated. For example, it may be desirable to improve the binding affinity and / or other biological properties of the antigen-binding molecules. Amino acid sequence variants of the antigen-binding molecules are produced by introducing appropriate nucleotide changes into the nucleic acid of the antigen-binding molecule or by peptide synthesis. All of the amino acid sequence modifications described below should result in antibody-binding molecules that still retain the desired biological activity (binding to CD20 and CD22 and CD3) of the unmodified parent molecule.

[0176] The term "amino acid" or "amino acid residue" typically refers to amino acids having an art-recognized definition, such as amino acids selected from the group consisting of alanine (Ala or A); arginine (Arg or R); asparagine (Asn or N); aspartic acid (Asp or D); cysteine ​​(Cys or C); glutamine (Gln or Q); glutamic acid (Glu or E); glycine (Gly or G); histidine (His or H); isoleucine (He or I); leucine (Leu or L); lysine (Lys or K); methionine (Met or M); phenylalanine (Phe or F); proline (Pro or P); serine (Ser or S); threonine (Thr or T); tryptophan (Trp or W); tyrosine (Tyr or Y); and valine (Val or V), although modified, synthetic, or rare amino acids may be used if desired. In general, amino acids can be classified according to the presence of a nonpolar side chain (e.g., Ala, Cys, He, Leu, Met, Phe, Pro, Val); a negatively charged side chain (e.g., Asp, Glu); a positively charged side chain (e.g., Arg, His, Lys); or an uncharged polar side chain (e.g., Asn, Cys, Gln, Gly, His, Met, Phe, Ser, Thr, Trp, and Tyr).

[0177] Amino acid modifications include, for example, deletion from and / or insertion into and / or substitution of residues in the amino acid sequence of the antigen-binding molecule. Any combination of deletion, insertion, and substitution can be performed to arrive at the final construct, provided that the final construct has the desired properties. Amino acid changes can also modify post-translational processes of the antigen-binding molecule, such as changing the number or position of glycosylation sites.

[0178] For example, 1, 2, 3, 4, 5, or 6 amino acids may be inserted, substituted, or deleted in each of the CDRs (depending, of course, on their length), while 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or 25 amino acids may be inserted, substituted, or deleted in each of the FRs. Preferably, the insertion of amino acid sequences into the antigen-binding molecule includes amino- and / or carboxyl-terminal fusions ranging in length from 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 residues to polypeptides containing 100 or more residues, as well as intrasequence insertions of single or multiple amino acid residues. Corresponding modifications may also be made in the third domain of the antigen-binding molecule of the present invention. Insertion variants of the antigen-binding molecule of the present invention include fusion of an enzyme to the N-terminus or C-terminus of the antigen-binding molecule, or fusion to a polypeptide.

[0179] The most important sites for substitution mutagenesis include (but are not limited to) the CDRs of the heavy and / or light chains, especially the hypervariable regions, but also contemplate modification of the FRs in the heavy and / or light chains. Substitutions are preferably conservative substitutions as described herein. Preferably, depending on the length of the CDR or FR, 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acids may be substituted in the CDRs, while 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or 25 amino acids may be substituted in the framework regions (FRs). For example, if the CDR sequence comprises 6 amino acids, it is contemplated that 1, 2, or 3 of these amino acids are substituted. Similarly, if the CDR sequence comprises 15 amino acids, it is contemplated that 1, 2, 3, 4, 5, or 6 of these amino acids are substituted.

[0180] A useful method for identifying certain residues or regions of an antigen-binding molecule that are preferred locations for mutagenesis is called "alanine scanning mutagenesis," described by Cunningham and Wells in Science, 244:1081-1085 (1989). In this method, a residue or target residues in an antigen-binding molecule are identified (e.g., charged residues such as arg, asp, his, lys, and glu) that affect the interaction of the amino acid with the epitope, and replaced with neutral or negatively charged amino acids (most preferably alanine or polyalanine).

[0181] Next, further or other variants are introduced at or in place of the substitution site to select those amino acid positions that show functional sensitivity to the substitution. Thus, the site or region where the amino acid sequence variant is introduced is predetermined, but the nature of the mutation itself does not need to be predetermined. For example, to analyze or optimize the performance of a mutation at a given site, alanine scanning or random mutagenesis may be performed at the target codon or target region, and the expressed antigen-binding molecule variants are screened for the optimal combination of desired activities. Techniques for making substitution mutations at a predetermined site in DNA with a known sequence are known, such as M13 primer mutagenesis and PCR mutagenesis. Screening of mutants is performed using an assay of antigen-binding activity, such as CD20 and CD22 or CD3 binding.

[0182] In general, when amino acids are replaced in one or more or all of the CDRs of the heavy and / or light chain, the resulting "replaced" sequence is preferably at least 60% or 65%, more preferably 70% or 75%, even more preferably 80% or 85%, and particularly preferably 90% or 95% identical to the "original" CDR sequence. This means that how identical it is to the "replaced" sequence depends on the length of the CDR. For example, a CDR with 5 amino acids is preferably 80% identical to the replaced sequence to have at least one replaced amino acid. Thus, the CDRs of an antigen-binding molecule may have different degrees of identity to their replacement sequences, for example, CDRL1 may have 80%, while CDRL3 may have 90%.

[0183] Preferred substitutions (or replacements) are conservative substitutions. However, any substitution (including non-conservative substitutions or one or more of the "exemplary substitutions" listed in Table 3 below) is envisaged, as long as the antigen binding molecule retains the ability to bind CD20 and CD22 via the first binding domain and to bind CD3 epsilon via the second binding domain, and / or its CDRs have identity to the replaced sequence (at least 60% or 65%, more preferably 70% or 75%, even more preferably 80% or 85%, particularly preferably 90% or 95% identical to the "original" CDR sequence).

[0184] Conservative substitutions are shown under the heading of "preferred substitutions" in Table 3. If such substitutions alter biological activity, then they are referred to in Table 3 as "exemplary substitutions," or more substantial changes, as further described below in connection with amino acid classes, can be introduced and the products screened for desired characteristics.

[0185] [Table 3]

[0186] Substantial alterations in the biological properties of the antigen-binding molecules of the present invention are achieved by selecting substitutions that differ significantly in their impact on (a) the structure of the polypeptide backbone in the substituted region, e.g., as a sheet or helical conformation, (b) the charge or hydrophobicity of the molecule at the target site, or (c) maintaining the bulk of the side chain. Naturally occurring residues are classified into the following groups based on common side chain properties: (1) hydrophobic: norleucine, met, ala, val, leu, ile; (2) neutral hydrophobic: cys, ser, thr; asn, gln (3) acidic: asp, glu; (4) basic: his, lys, arg; (5) residues that affect chain orientation: gly, pro; and (6) aromatic: trp, tyr, phe.

[0187] Non-conservative substitutions would involve exchanging a member of one of these classes for another. Any cysteine ​​residue not involved in maintaining the proper conformation of the antigen-binding molecule may be substituted, generally with serine, to improve the oxidative stability of the molecule and avoid aberrant cross-linking. Conversely, adding cysteine ​​bond(s) to an antibody may improve its stability, particularly if the antibody is an antibody fragment such as an Fv fragment.

[0188] For amino acid sequences, sequence identity and / or similarity are determined by standard techniques known in the art, such as, but not limited to, the local sequence identity algorithm of Smith and Waterman, 1981, Adv. Appl. Math. 2:482, the sequence identity alignment algorithm of Needleman and Wunsch, 1970, J. Mol. Biol. 48:443, the search for similarity method of Pearson and Lipman, 1988, Proc. Nat. Acad. Sci. USA 85:2444, computerized implementations of these algorithms (GAP, BESTFIT, FASTA and TFASTA in the Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Drive, Madison, Wis.), the Best Fit sequence program described by Devereux et al., 1984, Nucl. Acid Res. 12:387-395, preferably using the default settings, or by visual inspection, etc. Preferably, the percent identity is calculated by FastDB based on the following parameters: mismatch penalty of 1; gap penalty of 1; gap size penalty of 0.33; and joining penalty of 30, “Current Methods in Sequence Comparison and Analysis,” Macromolecule Sequencing and Synthesis, Selected Methods and Applications, pp 127-149 (1988), Alan R. Liss, Inc.

[0189] An example of a useful algorithm is PILEUP. PILEUP generates a multiple sequence alignment from a group of related sequences using a progressive pairwise alignment method. It can also plot a tree showing the clustering relationships used to generate the alignment. PILEUP uses a simplified version of the progressive alignment method of Feng & Doolittle, 1987, J. Mol. Evol. 35:351-360; this method is similar to that described in Higgins and Sharp, 1989, CABIOS 5:151-153. Useful PILEUP parameters include a default gap weight of 3.00, a default gap length weight of 0.10, and weighted end gaps.

[0190] Another example of a useful algorithm is the BLAST algorithm described in Altschul et al., 1990, J. Mol. Biol. 215:403-410; Altschul et al., 1997, Nucleic Acids Res. 25:3389-3402; and Karin et al., 1993, Proc. Natl. Acad. Sci. USA 90:5873-5787. A particularly useful BLAST program is the WU-BLAST-2 program, obtained from Altschul et al., 1996, Methods in Enzymology 266:460-480. WU-BLAST-2 uses several search parameters, most of which are set to default values. Adjustable parameters are set to the following values: overlap span=1, overlap fraction=0.125, word threshold (T)=II. The HSP S and HSP S2 parameters are dynamic values ​​and are established by the program itself depending on the composition of the particular sequence and the composition of the particular database in which the sequence of interest is to be searched; however, the values ​​can be adjusted to increase sensitivity.

[0191] An additional useful algorithm is Gapped BLAST, reported by Altschul et al., 1993, Nucl. Acids Res. 25:3389-3402. Gapped BLAST uses the BLOSUM-62 substitution score, the threshold T parameter is set to 9, the two-hit method which results in ungapped extensions has a cost of 10+k for a gap length k, Xu is set to 16, and Xg is set to 40 for the database search stage and 67 for the output stage of the algorithm. Gapped alignments are initiated by a score corresponding to approximately 22 bits.

[0192] Generally, the amino acid homology, similarity, or identity between the individual variant CDR or VH / VL sequences is at least 60% relative to the sequences shown herein, and more typically, the homology or identity is increased to at least 65% or 70%, more preferably at least 75% or 80%, and even more preferably at least 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, and nearly 100%. Similarly, "percent (%) nucleic acid sequence identity" with respect to the nucleic acid sequences of the binding proteins identified herein is defined as the percentage of nucleotide residues in a candidate sequence that are identical to nucleotide residues in the coding sequence of the antigen-binding molecule. A specific method utilizes the BLASTN module of WU-BLAST-2 set to the default parameters, with overlap span and overlap fraction set to 1 and 0.125, respectively.

[0193] Generally, the nucleic acid sequence homology, similarity or identity between the nucleotide sequence encoding each variant CDR or VH / VL sequence and the nucleotide sequences set out herein will be at least 60%, more typically the homology or identity will be at least 65%, 70%, 75%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% and preferably increasing to nearly 100%. Thus, a "variant CDR" or "variant VH / VL region" is one that has a particular homology, similarity or identity to a parent CDR / VH / VL of the invention and shares a biological function, including but not limited to, at least 60%, 65%, 70%, 75%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% of the specificity and / or activity of the parent CDR or VH / VL.

[0194] In one embodiment, the percentage of identity of the antigen-binding molecule of the present invention to human germline is ≧70% or ≧75%, more preferably ≧80% or ≧85%, even more preferably ≧90%, and most preferably ≧91%, ≧92%, ≧93%, ≧94%, ≧95% or even ≧96%. Identity to human antibody germline gene products is considered to be an important feature for reducing the risk of a therapeutic protein inducing an immune response against the drug in the treated patient. Hwang & Foote ("Immunogenicity of engineered antibodies"; Methods 36 (2005) 3-10) demonstrate that the reduction of the non-human portion of the drug-antigen-binding molecule results in a reduced risk of inducing anti-drug antibodies in the treated patient. By comparing a large number of clinically evaluated antibody drugs and their respective immunogenicity data, humanization of the V region of an antibody shows a tendency for the protein to be less immunogenic (5.1% of patients on average) than antibodies carrying unmodified non-human V regions (23.59% of patients on average). Therefore, for protein therapeutics in the form of antigen-binding molecules based on V regions, a high degree of identity to human sequences is desirable. To determine this germline identity, the V region of the VL can be aligned with the amino acid sequences of human germline V and J segments (http: / / vbase.mrc-cpe.cam.ac.uk / ) using Vector NTI software, and the amino acid sequence can be calculated in percentage by dividing the number of identical amino acid residues by the total number of amino acid residues of the VL. A similar method is possible for the VH segment (http: / / vbase.mrc-cpe.cam.ac.uk / ), with the exception that the VH CDR3 can be excluded due to its high diversity and lack of existing human germline VH CDR3 alignment partners. Recombinant techniques can then be used to increase sequence identity to human antibody germline genes.

[0195] In a further embodiment, the bispecific antigen-binding molecules of the invention exhibit high monomer yields under standard laboratory-scale conditions (e.g., a standard two-step purification process). Preferably, the monomer yield of the antigen-binding molecules of the invention is ≧0.25 mg / L of supernatant, more preferably ≧0.5 mg / L, even more preferably ≧1 mg / L, and most preferably ≧3 mg / L of supernatant.

[0196] Similarly, the yield of dimeric antigen-binding molecule isoforms, and therefore the percentage of monomers in the antigen-binding molecule (i.e., monomer: (monomer + dimer)), can be determined. The productivity of monomeric and dimeric antigen-binding molecules, as well as the calculated percentage of monomers, can be obtained, for example, by an SEC purification step of culture supernatants derived from standardized research-scale production in roller bottles. In one embodiment, the percentage of monomers in the antigen-binding molecule is ≧80%, more preferably ≧85%, even more preferably ≧90%, and most preferably ≧95%.

[0197] In one embodiment, the antigen-binding molecule preferably has a plasma stability (ratio of EC50 in the presence of plasma to EC50 in the absence of plasma) of ≦5 or ≦4, more preferably ≦3.5 or ≦3, even more preferably ≦2.5 or ≦2, and most preferably ≦1.5 or ≦1. The plasma stability of the antigen-binding molecule is determined by incubating the construct in human plasma at 37° C. for 24 hours, followed by ELISA. 51The antigen-binding molecule may be tested by determining the EC50 in a chromium release cytotoxicity assay. The effector cells in the cytotoxicity assay may be stimulated enriched human CD8 positive T cells. The target cells may be, for example, CHO cells transfected with human CD20 and CD22. The effector cell to target cell (E:T) ratio may be selected as 10:1 or 5:1. The human plasma pool used for this purpose is derived from blood of healthy donors collected by EDTA-coated syringes. The cellular components are removed by centrifugation, and the upper plasma phase is collected and then pooled. As a control, the antigen-binding molecule is diluted in RPMI-1640 medium immediately before the cytotoxicity assay. The plasma stability is calculated as the ratio of EC50 (after plasma incubation) to EC50 (control).

[0198] It is further preferred that the conversion rate of the antigen-binding molecule of the present invention from monomer to dimer is low. The conversion rate can be measured under different conditions and analyzed by high-performance size exclusion chromatography. For example, the incubation of the monomeric isoform of the antigen-binding molecule can be performed in an incubator at a concentration of, for example, 100 μg / ml or 250 μg / ml and at 37° C. for 7 days. Under these conditions, it is preferred that the antigen-binding molecule of the present invention exhibits a percentage of dimer of ≦5%, more preferably ≦4%, even more preferably ≦3%, even more preferably ≦2.5%, even more preferably ≦2%, even more preferably ≦1.5%, and most preferably ≦1% or ≦0.5% or even 0%.

[0199] It is also preferred that the bispecific antigen-binding molecule of the present invention exhibits a very low dimer conversion rate after several freeze / thaw cycles. For example, the antigen-binding molecule monomer is adjusted to a concentration of 250 μg / ml, for example, in a universal formulation buffer, and subjected to three freeze / thaw cycles (freezing at -80°C for 30 minutes, then thawing at room temperature for 30 minutes), followed by high-speed SEC to determine the percentage of the initial monomeric antigen-binding molecule converted to a dimeric antigen-binding molecule. Preferably, the percentage of dimers in the bispecific antigen-binding molecule is, for example, ≦5% after three freeze / thaw cycles, more preferably ≦4%, even more preferably ≦3%, even more preferably ≦2.5%, even more preferably ≦2%, even more preferably ≦1.5%, and most preferably ≦1% or even ≦0.5%.

[0200] The bispecific antigen-binding molecules of the present invention preferably exhibit good thermal stability with an aggregation temperature of ≧45° C. or ≧50° C., more preferably ≧52° C. or ≧54° C., even more preferably ≧56° C. or ≧57° C., and most preferably ≧58° C. or ≧59° C. In terms of the aggregation temperature of the antibody, the thermal stability parameter can be determined as follows: An antibody solution at a concentration of 250 μg / ml is transferred into a single-use cuvette and placed in a dynamic light scattering (DLS) instrument. The sample is heated from 40° C. to 70° C. at a heating rate of 0.5° C. / min while the measurement radius is constantly acquired. The increase in radius, which indicates the melting and aggregation of the protein, is used to calculate the aggregation temperature of the antibody.

[0201] Alternatively, melting temperature curves can be determined by differential scanning calorimetry (DSC) to determine the intrinsic biophysical protein stability of antigen-binding molecules. These experiments are performed using a MicroCal LLC (Northampton, MA, USA) VP-DSC instrument. The energy uptake of samples containing antigen-binding molecules is recorded from 20°C to 90°C and compared to samples containing only formulation buffer. The antigen-binding molecules are adjusted to a final concentration of, for example, 250 μg / ml in SEC running buffer. The temperature of the entire sample is increased stepwise to record each melting curve. The energy uptake of the samples and formulation buffer standard at each temperature T is recorded. The difference in energy uptake Cp (kcal / mole / °C) of the samples minus the standard is plotted against each temperature. The melting temperature is defined as the temperature at which the energy uptake first reaches a maximum.

[0202] The CD20 and CD22xCD3 bispecific antigen-binding molecules of the target cells of the present invention are also envisaged to have a turbidity (measured by OD340 after concentrating purified monomeric antigen-binding molecules to 2.5 mg / ml and incubating overnight) of ≦0.2, preferably ≦0.15, more preferably ≦0.12, even more preferably ≦0.1, and most preferably ≦0.08.

[0203] In a further embodiment, the antigen-binding molecule according to the present invention is stable at physiological pH or slightly lower pH (i.e., about pH 7.4 to 6.0). The higher the resistance of the antigen-binding molecule at non-physiological pH (e.g., about pH 6.0), the higher the recovery rate of the antigen-binding molecule eluted from the ion exchange column relative to the total amount of protein loaded. The recovery rate of the antigen-binding molecule from the ion (e.g., cation) exchange column at about pH 6.0 is preferably ≧30%, more preferably ≧40%, more preferably ≧50%, even more preferably ≧60%, even more preferably ≧70%, even more preferably ≧80%, even more preferably ≧90%, even more preferably ≧95%, and most preferably ≧99%.

[0204] It is further envisaged that the bispecific antigen-binding molecules of the invention exhibit therapeutic efficacy or anti-tumour activity, which may be assessed, for example, in the test disclosed in the generalised examples below in advanced stage human tumour xenograft models.

[0205] On day 1 of the study, 5 × 10 6 Cells are injected subcutaneously into the right dorsal flank of female NOD / SCID mice. The average tumor volume is approximately 100 mm 3 When the total number of in vitro expanded human CD3 positive T cells reaches approximately 2 x 10 7 Mice are implanted by injection of 1000 cells. Mice in vehicle control group 1 do not receive effector cells and are used as non-implanted controls for comparison with vehicle control group 2 (receiving effector cells) to monitor the effect of T cells alone on tumor growth. The mean tumor volume is approximately 200 mm 3 Antibody treatment is initiated when tumor size reaches 100%. The mean tumor size of each treatment group on the treatment initiation day should not be statistically different from any other group (analysis of variance). Mice are treated with CD20 and CD22xCD3 bispecific antigen-binding molecules at 0.5 mg / kg / day by intravenous bolus injection for approximately 15-20 days. Tumors are measured by caliper during the study and progression is assessed by between-group comparison of tumor volume (TV). Tumor growth inhibition T / C [%] is determined by calculating TV as T / C% = 100 x (median TV of analyzed group) / (median TV of control group 2).

[0206] Those skilled in the art know how to vary or adapt the specific parameters of this test, such as the number of tumor cells injected, the injection site, the number of human T cells implanted, the amount of bispecific antigen-binding molecule to be administered, and the timeline, while still obtaining meaningful and reproducible results. Preferably, the tumor growth inhibition T / C[%] is ≦70 or ≦60, more preferably ≦50 or ≦40, even more preferably ≦30 or ≦20, and most preferably ≦10 or ≦5 or even ≦2.5. Preferably, the tumor growth inhibition is close to 100%.

[0207] In a preferred embodiment of the antigen-binding molecule of the present invention, the antigen-binding polypeptide is a single-chain antigen-binding molecule.

[0208] In a preferred embodiment of the antigen-binding molecule of the present invention, the third domain is composed of, in order from amino to carboxyl, It comprises hinge-CH2-CH3-linker-hinge-CH2-CH3.

[0209] In one embodiment of the invention, each of the polypeptide monomers of the third domain has an amino acid sequence that is at least 90% identical to a sequence selected from the group consisting of SEQ ID NOs: 17 to 24. In a preferred embodiment or invention, each of the polypeptide monomers has an amino acid sequence selected from SEQ ID NOs: 17 to 24.

[0210] Also, in one embodiment of the present invention, the CH2 domain of one or preferably each (both) polypeptide monomer of the third domain comprises an intradomain cysteine ​​disulfide bridge. As known in the art, the term "cysteine ​​disulfide bridge" refers to a functional group having the general structure RSSR. This linkage, also called SS bond or disulfide bridge, is obtained by coupling of two thiol groups of cysteine ​​residues. With respect to the antigen-binding molecules of the present invention, it is particularly preferred that the cysteines that form the cysteine ​​disulfide bridge in the mature antigen-binding molecule are introduced into the amino acid sequence of the CH2 domain corresponding to 309 and 321 (Kabat numbering).

[0211] In one embodiment of the invention, the glycosylation site at Kabat position 314 of the CH2 domain is removed. This removal of glycosylation site is preferably achieved by an N314X substitution, where X is any amino acid other than Q. Said substitution is preferably N314G. In a more preferred embodiment, said CH2 domain further comprises the following substitutions (positions according to Kabat): V321C and R309C (these substitutions introduce intradomain cysteine ​​disulfide bridges at Kabat positions 309 and 321).

[0212] For example, the preferred features of the antigen-binding molecules of the present invention compared to bispecific hetero-Fc antigen-binding molecules known in the art (Figure F 1b) may be related, inter alia, to the introduction of the above-mentioned modifications in the CH2 domain. Thus, with respect to the constructs of the present invention, it is preferred that the CH2 domain in the third domain of the antigen-binding molecules of the present invention comprises intradomain cysteine ​​disulfide bridges at Kabat positions 309 and 321, and / or the glycosylation site at Kabat position 314 is removed, preferably by N314G substitution.

[0213] In a further preferred embodiment of the present invention, the CH2 domain in the third domain of the antigen-binding molecule of the present invention comprises intradomain cysteine ​​disulfide bridges at Kabat positions 309 and 321, and the glycosylation site at Kabat position 314 is removed by an N314G substitution. Most preferably, the polypeptide monomer of the third domain of the antigen-binding molecule of the present invention has an amino acid sequence selected from the group consisting of SEQ ID NOs: 17 and 18.

[0214] In one embodiment, the present invention provides an antigen binding molecule comprising: (i) the first domain comprises two antibody variable domains and the second domain comprises two antibody variable domains; (ii) the first domain comprises one antibody variable domain and the second domain comprises two antibody variable domains; (iii) the first domain comprises two antibody variable domains and the second domain comprises one antibody variable domain; or (iv) the first domain comprises one antibody variable domain and the second domain comprises one antibody variable domain. An antigen-binding molecule is provided.

[0215] Thus, the first domain and the second domain may each be a binding domain comprising two antibody variable domains, such as a VH and a VL domain. Examples of such binding domains comprising two antibody variable domains as described herein above include, for example, Fv fragments, scFv fragments or Fab fragments as described herein above. Alternatively, either or both of the binding domains may comprise only a single variable domain. Examples of such single domain binding domains as described herein above include, for example, nanobodies or single variable domain antibodies comprising only one variable domain, which may be a VHH, VH or VL that specifically binds to an antigen or epitope independently of other V regions or domains.

[0216] In a preferred embodiment of the antigen-binding molecule of the present invention, the first and second domains are fused to the third domain via a peptide linker. Preferred peptide linkers are described herein above and are characterized by the amino acid sequence Gly-Gly-Gly-Gly-Ser (i.e., Gly4Ser (SEQ ID NO: 1)), or a polymer thereof (i.e., (Gly4Ser)x), where x is an integer equal to or greater than 1 (e.g., 2 or 3). A particularly preferred linker for fusing the first and second domains to the third domain is shown in SEQ ID NO: 1.

[0217] In a preferred embodiment, the antigen-binding molecule of the present invention has the following amino acid sequence, in order from amino to carboxyl: (a) the first domain; (b) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 1 to 3; (c) the second domain; (d) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 1, 2, 3, 9, 10, 11, and 12; (e) a first polypeptide monomer of the third domain; (f) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 5, 6, 7, and 8; and (g) a second polypeptide monomer of the third domain. The present invention is characterized by comprising:

[0218] The antigen-binding molecule of the present invention comprises a first domain that binds to CD20 and CD22, preferably to the extracellular domain (ECD) of CD20 and CD22. In the context of the present invention, the term "binding to the extracellular domain of CD20 and CD22" is understood to mean that the binding domain binds to CD20 and CD22 expressed on the surface of a target cell. Thus, the first domain according to the present invention preferably binds to CD20 and CD22 when expressed by a cell or cell line that naturally expresses CD20 and CD22 and / or when expressed by a cell or cell line that has been transformed or (stable / transiently) transfected with CD20 and CD22. In a preferred embodiment, the first binding domain also binds to CD20 and CD22 when CD20 and CD22 are used as "target" or "ligand" molecules in in vitro binding assays such as BIAcore or Scatchard. A "target cell" can be any prokaryotic or eukaryotic cell that expresses CD20 and CD22 on its surface; preferably, the target cell is a cell that is part of a human or animal body, such as a specific CD20- and CD22-expressing cancer or tumor cell.

[0219] Preferably, the first binding domain binds to human CD20 and CD22 / CD20 and CD22 ECD. In a more preferred embodiment, it binds to macaque CD20 and CD22 / CD20 and CD22 ECD. According to a most preferred embodiment, it binds to both human and macaque CD20 and CD22 / CD20 and CD22 ECD. "CD20 and CD22 extracellular domain" or "CD20 and CD22 ECD" refers to a region or sequence of CD20 and CD22 that is essentially free of the transmembrane and cytoplasmic domains of CD20 and CD22. It will be understood by those skilled in the art that the transmembrane domains identified for the CD20 and CD22 polypeptides of the present invention are identified according to criteria routinely used in the art for identifying hydrophobic domains of that type. The exact boundaries of the transmembrane domains may vary, but are most likely no more than about 5 amino acids on either end of the domains specifically mentioned herein.

[0220] Preferred binding domains that bind to CD3 are disclosed in WO 2010 / 037836 and WO 2011 / 121110. Any binding domain for CD3 described in these applications may be used in the context of the present invention, however, preferred are the third binding domains having SEQ ID NO: 400 or 409 disclosed herein. SEQ ID NO: 409 is highly preferred.

[0221] The present invention further provides a polynucleotide / nucleic acid molecule encoding the antigen-binding molecule of the present invention. A polynucleotide is a biopolymer composed of 13 or more nucleotide monomers covalently linked in a chain. DNA (e.g., cDNA) and RNA (e.g., mRNA) are examples of polynucleotides with different biological functions. Nucleotides are organic molecules that function as monomers or subunits of nucleic acid molecules such as DNA or RNA. A nucleic acid molecule or polynucleotide can be double-stranded and single-stranded, linear and circular. It is preferably contained within a vector that is contained within a host cell. The host cell can express the antigen-binding molecule, for example, after being transformed or transfected with a vector or polynucleotide of the present invention. For that purpose, the polynucleotide or nucleic acid molecule is operably linked to a control sequence.

[0222] The genetic code is the set of rules that translate the information coded in genetic material (nucleic acids) into proteins. Biological decoding in living cells is performed by ribosomes, which carry amino acids and attach them in the order specified by the mRNA using tRNA molecules that read the mRNA three nucleotides at a time. This code defines how triplet nucleotide sequences called codons specify the amino acid that will be added next during protein synthesis. With some exceptions, a three-nucleotide codon in a nucleic acid sequence specifies one amino acid. Because most genes are coded with the exact same code, this particular code is often referred to as the canonical or standard genetic code. Although the genetic code determines the protein sequence of a given coding region, other genomic regions can influence when and where these proteins are produced.

[0223] Furthermore, the present invention provides vectors comprising the polynucleotide / nucleic acid molecules of the present invention. A vector is a nucleic acid molecule used as a vehicle to transfer (foreign) genetic material into a cell. The term "vector" includes, but is not limited to, plasmids, viruses, cosmids, and artificial chromosomes. In general, genetically engineered vectors contain an origin of replication, a multiple cloning site, and a selection marker. The vector itself is generally a nucleotide sequence (generally a DNA sequence) that contains an insert (transgene) and a larger sequence that serves as the "backbone" of the vector. Modern vectors may contain additional features in addition to the transgene insert and backbone: promoters, genetic markers, antibiotic resistance, reporter genes, targeting sequences, protein purification tags. Vectors called expression vectors (expression constructs) are specifically intended for the expression of a transgene in a target cell and generally have regulatory sequences.

[0224] The term "control sequence" refers to a DNA sequence necessary for the expression of an operably linked coding sequence in a particular host organism. Suitable control sequences for prokaryotes include, for example, a promoter, an optional operator sequence, and a ribosome binding site. Eukaryotic cells are known to utilize promoters, polyadenylation signals, and enhancers.

[0225] A nucleic acid is "operably linked" when it is placed into a functional relationship with another nucleic acid sequence. For example, DNA for a presequence or secretory leader is operably linked to DNA for a polypeptide if it is expressed as a protein precursor that participates in the secretion of the polypeptide; a promoter or enhancer is operably linked to a coding sequence if it affects the transcription of the sequence; or a ribosome binding site is operably linked to a coding sequence if it is positioned so as to facilitate translation. Generally, "operably linked" means that the DNA sequences being linked are contiguous, and, in the case of a secretory leader, contiguous and in reading frame. Enhancers, however, need not be contiguous. Linking is accomplished by ligation at convenient restriction sites. If such sites do not exist, synthetic oligonucleotide adaptors or linkers are used in accordance with conventional practice.

[0226] "Transfection" is the process of deliberately introducing a nucleic acid molecule or polynucleotide (including a vector) into a target cell. The term is primarily used for non-viral methods in eukaryotic cells. Transduction is often used to describe the viral-mediated transfer of a nucleic acid molecule or polynucleotide. Transfection of animal cells typically involves creating a transient pore or "hole" in the cell membrane to allow uptake of the material. Transfection can be performed using calcium phosphate, by electroporation, by cell compression, or by mixing cationic lipids with a substance to create liposomes, which fuse with the cell membrane and accumulate the cargo inside.

[0227] The term "transformation" is used to describe the non-viral transfer of nucleic acid molecules or polynucleotides (including vectors) into bacteria and into non-animal eukaryotic cells, including plant cells. Transformation is thus the genetic modification of bacteria or non-animal eukaryotic cells resulting from the direct uptake through the cell membrane from its surroundings and subsequent incorporation of exogenous genetic material (nucleic acid molecules). Transformation can occur by artificial means. For transformation to occur, cells or bacteria must be in a competent state in which transformation can occur as a timed response to environmental conditions such as starvation and cell density.

[0228] Furthermore, the present invention provides a host cell transformed or transfected with a polynucleotide / nucleic acid molecule or vector of the present invention. As used herein, the term "host cell" or "recipient cell" is intended to include any individual cell or cell culture that can be or has been a recipient of vectors, exogenous nucleic acid molecules, and polynucleotides encoding the antigen-binding molecules of the present invention; and / or the antigen-binding molecule itself. Introduction of the respective substance into the cell is performed by transformation, transfection, etc. The term "host cell" is also intended to include the progeny or potential progeny of a single cell. Since certain modifications may occur in successive generations, either due to spontaneous, accidental, or deliberate mutations, or due to environmental influences, such progeny may not in fact be completely identical (morphologically, or in terms of genome or total DNA set) to the parent cell, but still fall within the scope of the term as used herein. Suitable host cells include prokaryotic or eukaryotic cells, and include, but are not limited to, bacteria, yeast cells, fungal cells, plant cells, and animal cells, such as insect cells and mammalian cells, such as mouse, rat, macaque, or human cells.

[0229] The antigen-binding molecules of the present invention can be produced in bacteria. After expression, the antigen-binding molecules of the present invention can be isolated from the E. coli cell paste in a soluble fraction and purified, for example, by affinity chromatography and / or size exclusion. Final purification can be carried out, for example, in the same manner as the purification method of antibodies expressed in CHO cells.

[0230] In addition to prokaryotes, eukaryotic microbes such as filamentous fungi or yeast are suitable cloning or expression hosts for the antigen-binding molecules of the present invention. Saccharomyces cerevisiae, or common baker's yeast, is the most commonly used among lower eukaryotic host microorganisms. However, several other genera, species, and strains are commonly available and useful herein, such as Schizosaccharomyces pombe, Kluyveromyces hosts, such as K. lactis, K. fragilis (ATCC 12424), K. bulgaricus (ATCC 16045), K. wickeramii (ATCC 24178), K. waltii (ATCC 56500), K. drosophilarum (ATCC 56661), K. spp. (ATCC 56662), K. spp. (ATCC 56663), K. spp. (ATCC 56664), K. spp. (ATCC 56665), K. spp. (ATCC 56666), K. spp. (ATCC 56667), K. spp. (ATCC 56668), K. spp. (ATCC 56669), K. spp. (ATCC 56669), K. spp. (ATCC 56669), K. spp. (ATCC 56661), K. spp. (ATCC 56662), K. spp. (ATCC 56663), K. spp. (ATCC 56664), K. spp. (ATCC 56666), K. spp. (ATCC 56667), K. spp. (ATCC 56668), K. spp. (ATCC 56669 ... 36906, K. thermotolerans, and K. marxianus; Yarrowia (EP 402226); Pichia pastoris (EP 183070); Candida; Trichoderma reesia (EP 244234); Neurospora crassa; Schwanniomyces, for example Schwanniomyces occidentalis. occidentalis; and filamentous fungi such as Neurospora, Penicillium, Tolypocladium, and Aspergillus hosts such as A. nidulans and A. niger.

[0231] Suitable host cells for the expression of the glycosylated antigen-binding molecules of the present invention are derived from multicellular organisms. Examples of invertebrate cells include plant cells and insect cells. Many strains and variants of baculoviruses and corresponding permissive insect host cells from hosts (e.g., Spodoptera frugiperda (caterpillar), Aedes aegypti (mosquito), Aedes albopictus (mosquito), Drosophila melanogaster (fruit fly), and Bombyx mori) have been identified. Various virus strains for transfection are publicly available, such as the L-1 variant of Autographa californica NPV and the Bm-5 strain of Bombyx mori NPV, and such viruses may be used as viruses herein according to the invention, particularly for transfection of Spodoptera frugiperda cells.

[0232] Plant cell cultures of cotton, corn, potato, soybean, petunia, tomato, Arabidopsis and tobacco can also be used as hosts. Cloning and expression vectors useful for producing proteins in plant cell cultures are known to those skilled in the art. See, for example, Hiatt et al., Nature (1989) 342:76-78; Owen et al. (1992) Bio / Technology 10:790-794; Artsaenko et al. (1995) The Plant J 8:745-750; and Fecker et al. (1996) Plant Mol Biol 32:979-986.

[0233] However, interest has been greatest in vertebrate cells, and propagation of vertebrate cells in culture (tissue culture) has become routine procedure. Examples of useful mammalian host cell lines are: monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); human embryonic kidney line (293 cells, or 293 cells subcloned for growth in suspension culture, Graham et al., J. Gen Virol. 36:59 (1977)); baby hamster kidney cells (BHK, ATCC CCL 10); Chinese hamster ovary cells / -DHFR (CHO, Urlaub et al., Proc. Natl. Acad. Sci. USA 77:4216 (1980)); mouse Sertoli cells (TM4, Mather, Biol. Reprod. 23:243-251 (1980)); monkey kidney cells (CVI ATCC CCL 70); African green monkey kidney cells (VERO-76, ATCC CRL 1587); human cervical carcinoma cells (HELA, ATCC CCL 2); canine kidney cells (MDCK, ATCC CCL 34); buffalo rat liver cells (BRL 3A, ATCC CRL 1442); human lung cells (W138, ATCC CCL 75); human liver cells (Hep G2, 1413 8065); mouse mammary tumor (MMT 060562, ATCC CCL5 1); TRI cells (Mather et al., Annals NY Acad. Sci. (1982) 383:44-68); MRC 5 cells; FS4 cells; and a human hepatoma line (Hep G2).

[0234] In a further embodiment, the present invention provides a process for producing an antigen-binding molecule of the present invention, the process comprising culturing a host cell of the present invention under conditions allowing expression of the antigen-binding molecule of the present invention, and recovering the produced antigen-binding molecule from the culture.

[0235] As used herein, the term "culture" refers to the maintenance, differentiation, growth, proliferation, and / or propagation of cells in vitro under suitable conditions in a medium. The term "expression" includes any step involved in the production of an antigen-binding molecule of the present invention, including, but not limited to, transcription, post-transcriptional modification, translation, post-translational modification, and secretion.

[0236] When using recombinant techniques, antigen-binding molecules can be produced in the periplasmic space intracellularly or directly secreted into the medium. If the antigen-binding molecule is produced intracellularly, as a first step, particulate debris, host cells or lysed fragments are removed, for example, by centrifugation or ultrafiltration. Carter et al., Bio / Technology 10:163-167 (1992) describes a procedure for isolating antibodies secreted into the periplasmic space of Escherichia coli (E. coli). Briefly, cell paste is thawed in the presence of sodium acetate (pH 3.5), EDTA, and phenylmethylsulfonyl fluoride (PMSF) for about 30 minutes. Cell debris can be removed by centrifugation. If the antibody is secreted into the medium, the supernatant from such an expression system is generally first concentrated using a commercially available protein concentration filter (e.g., an Amicon or Millipore Pellicon ultrafiltration unit). A protease inhibitor such as PMSF may be included in any of the foregoing steps to inhibit proteolysis, and antibiotics may be included to prevent the growth of adventitious contaminating bacteria.

[0237] The antigen-binding molecule of the present invention prepared from the host cell can be recovered or purified using, for example, hydroxyapatite chromatography, gel electrophoresis, dialysis, and affinity chromatography. Depending on the antibody to be recovered, other protein purification techniques, such as fractionation on an ion exchange column, ethanol precipitation, reverse phase HPLC, chromatography on silica, chromatography on heparin SEPHAROSE™, chromatography on anion or cation exchange resin (e.g., polyaspartic acid column), chromatofocusing, SDS-PAGE, and ammonium sulfate precipitation, can also be used. When the antigen-binding molecule of the present invention contains a CH3 domain, Bakerbond ABX resin (JT Baker, Phillipsburg, NJ) is useful for purification.

[0238] Affinity chromatography is the preferred purification technique. The matrix to which the affinity ligand is attached is most often agarose, although other matrices are available. Mechanically stable matrices (e.g., controlled pore glass or poly(styrenedivinyl)benzene) allow for faster flow rates and shorter processing times than can be achieved with agarose.

[0239] Furthermore, the present invention provides a pharmaceutical composition comprising the antigen-binding molecule of the present invention or an antigen-binding molecule produced according to the method of the present invention. In the pharmaceutical composition of the present invention, the homogeneity of the antigen-binding molecule is preferably ≧80%, more preferably ≧81%, ≧82%, ≧83%, ≧84%, or ≧85%, even more preferably ≧86%, ≧87%, ≧88%, ≧89%, or ≧90%, even more preferably ≧91%, ≧92%, ≧93%, ≧94%, or ≧95%, and most preferably ≧96%, ≧97%, ≧98%, or ≧99%.

[0240] As used herein, the term "pharmaceutical composition" relates to a composition suitable for administration to a patient (preferably a human patient). Particularly preferred pharmaceutical compositions of the present invention comprise one or more antigen-binding molecules of the present invention, preferably in a therapeutically effective amount. Preferably, the pharmaceutical composition further comprises a suitable formulation of one or more (pharmaceutical effective) carriers, stabilizers, excipients, diluents, solubilizers, surfactants, emulsifiers, preservatives, and / or adjuvants. Components of acceptable compositions are preferably non-toxic to recipients at the dosages and concentrations employed. Pharmaceutical compositions of the present invention include, but are not limited to, liquid compositions, frozen compositions, and lyophilized compositions.

[0241] The composition of the present invention may include a pharmaceutically acceptable carrier.Generally, as used herein, "pharmaceutically acceptable carrier" refers to any aqueous and non-aqueous solution, sterile solution, solvent, buffer solution, such as phosphate buffered saline (PBS) solution, water, suspension, emulsion such as oil / water emulsion, various types of wetting agents, liposome, dispersion medium, and coating that are suitable for pharmaceutical administration, especially parenteral administration.The use of such media and agents in pharmaceutical compositions is known in the art, and compositions containing such carriers can be formulated by known conventional methods.

[0242] Certain embodiments provide pharmaceutical compositions comprising an antigen-binding molecule of the invention and one or more excipients, such as those exemplarily described in this section and elsewhere herein. Excipients may be used in the invention for a wide range of purposes, such as adjusting the physical, chemical, or biological properties of the formulation, such as adjusting the viscosity, and / or improving the efficacy and / or stabilizing such formulations, as well as methods of the invention against deterioration and damage due to stresses that occur during and after manufacture, transportation, storage, preparation prior to use, and administration.

[0243] In certain embodiments, pharmaceutical compositions may contain formulation materials intended to, for example, modify, sustain or protect the pH, osmolality, viscosity, clarity, color, isotonicity, odor, sterility, stability, dissolution or release rate, adsorption or permeation of the composition (see REMINGTON'S PHARMACEUTICAL SCIENCES, 18" Edition, (ARGenrmo, ed.), 1990, Mack Publishing Company). In such embodiments, suitable formulation materials may include, but are not limited to, the following: Amino acids, such as glycine, alanine, glutamine, asparagine, threonine, proline, 2-phenylalanine, such as charged amino acids, preferably lysine, lysine acetate, arginine, glutamate, and / or histidine Antimicrobials, such as antibacterial and antifungal agents · Antioxidants such as ascorbic acid, methionine, sodium sulfite, or sodium bisulfite; Buffers, buffer systems, and buffering agents used to maintain the composition at or slightly below physiological pH, preferably at a lower pH of 4.0-6.5; examples of buffers are borate, bicarbonate, Tris-HCl, citrate, phosphate, or other organic acids, succinate, phosphate, and histidine; for example, Tris buffer at about pH 7.0-8.5; non-aqueous solvents, for example, propylene glycol, polyethylene glycol, vegetable oils such as olive oil, and injectable organic esters such as ethyl oleate; Aqueous carriers such as water, alcoholic / aqueous solutions, emulsions, or suspensions, e.g. saline and buffered media; · Biodegradable polymers such as polyester; Bulking agents such as mannitol or glycine; · Chelating agents such as ethylenediaminetetraacetic acid (EDTA); · Isotonic and absorption retarding agents; Complexing agents, such as caffeine, polyvinylpyrrolidone, β-cyclodextrin, or hydroxypropyl-β-cyclodextrin Injectables; monosaccharides; disaccharides; and other carbohydrates (e.g., glucose, mannose, or dextrins); the carbohydrates may be non-reducing sugars, preferably trehalose, sucrose, octasulfate, sorbitol, or xylitol; (low molecular weight) proteins, polypeptides or proteinaceous carriers, such as human or bovine serum albumin, gelatin or immunoglobulins, preferably of human origin; Colouring and flavouring agents; Sulfur-containing reducing agents, such as glutathione, thioctic acid, sodium thioglycolate, thioglycerol, [alpha]-monothioglycerol, and sodium thiosulfate Diluents; Emulsifiers; · Hydrophilic polymers such as polyvinylpyrrolidone; · Salt-forming counterions such as sodium; Preservatives, such as antimicrobials, antioxidants, chelating agents, inert gases and the like; examples are benzalkonium chloride, benzoic acid, salicylic acid, thimerosal, phenethyl alcohol, methylparaben, propylparaben, chlorhexidine, sorbic acid, or hydrogen peroxide; Metal complexes such as Zn-protein complexes; Solvents and co-solvents (e.g., glycerin, propylene glycol, or polyethylene glycol); sugars and sugar alcohols, such as trehalose, sucrose, octasulfate, mannitol, sorbitol or xylitol, stachyose, mannose, sorbose, xylose, ribose, myoinisitose, galactose, lactitol, ribitol, myoinisitol, galactitol, glycerol, cyclitols (e.g. inositol), polyethylene glycol; and polyhydric sugar alcohols; · Suspending agents; Surfactants or wetting agents, such as pluronics, PEG, sorbitan esters, polysorbates, such as polysorbate 20, polysorbate, triton, tromethamine, lecithin, cholesterol, tyloxapal; surfactants can be detergents, preferably with a molecular weight of >1.2 KD, and / or polyethers, preferably with a molecular weight of >3 KD; non-limiting examples of preferred detergents are Tween 20, Tween 40, Tween 60, Tween 80, and Tween 85; non-limiting examples of preferred polyethers are PEG 3000, PEG 3350, PEG 4000, and PEG 5000; · A stability enhancer such as sucrose or sorbitol; isotonicity enhancing agents, for example alkali metal halides, preferably sodium chloride or potassium chloride, mannitol, sorbitol; Parenteral delivery vehicles, such as sodium chloride solution, Ringer's dextrose, dextrose and sodium chloride, lactated Ringer's, or fixed oils; Intravenous delivery vehicles, such as fluid and nutrient replenishers, electrolyte replenishers (such as those based on Ringer's dextrose).

[0244] In the context of the present invention, the pharmaceutical composition, which may preferably be a liquid composition, or may be a solid composition obtained by lyophilization, or may be a reconstituted liquid composition, comprises: (a) an antigen-binding molecule comprising at least three domains, the first domain binds to a surface antigen of a target cell and has an isoelectric point (pI) in the range of 4 to 9.5; the second domain binds to a second antigen and has a pI in the range of 8 to 10, preferably 8.5 to 9.0; and the optional third domain comprises two polypeptide monomers, each of which comprises a hinge, a CH2 domain, and a CH3 domain, said two polypeptide monomers fused to each other via a peptide linker; (b) at least one buffering agent; (c) at least one sugar; and (d) at least one surfactant Including, The pH of the pharmaceutical composition is in the range of 3.5-6.

[0245] It is further envisaged in the context of the present invention that the at least one buffering agent is present in a concentration range of 5-200 mM, more preferably in a concentration range of 10-50 mM. It is envisaged in the context of the present invention that the at least one sugar is selected from the group consisting of monosaccharides, disaccharides, cyclic polysaccharides, sugar alcohols, linear branched dextrans or linear unbranched dextrans. It is also envisaged in the context of the present invention that the disaccharide is selected from the group consisting of sucrose, trehalose, and mannitol, sorbitol and combinations thereof. It is further envisaged in the context of the present invention that the sugar alcohol is sorbitol. It is envisaged in the context of the present invention that the at least one sugar is present in a concentration range of 1-15% (m / V), more preferably in a concentration range of 9-12% (m / V).

[0246] It is also envisaged in the context of the present invention that the at least one surfactant is selected from the group consisting of polysorbate 20, polysorbate 40, polysorbate 60, polysorbate 80, poloxamer 188, pluronic F68, triton X-100, polyoxyethylene, PEG 3350, PEG 4000, and combinations thereof. It is further envisaged in the context of the present invention that the at least one surfactant is present in a concentration ranging from 0.004 to 0.5% (m / V), preferably ranging from 0.001 to 0.01% (m / V). It is envisaged in the context of the present invention that the pH of the composition is in the range of 4.0 to 5.0, preferably 4.2. It is also envisaged in the context of the present invention that the pharmaceutical composition has an osmolality ranging from 150 to 500 mOsm. It is further envisaged in the context of the present invention that the pharmaceutical composition further comprises an excipient selected from the group consisting of one or more polyols and one or more amino acids. It is envisaged in the context of the present invention that the one or more excipients are present in a concentration range of 0.1-15% (w / V).

[0247] The pharmaceutical composition comprises (a) an antigen-binding molecule as discussed above; (b) 10 mM glutamate or acetate; (c) 9% (m / V) sucrose or 6% (m / V) sucrose and 6% (m / V) hydroxypropyl-β-cyclodextrin; (d) 0.01% (m / V) polysorbate 80 Including, The pH of the liquid pharmaceutical composition is 4.2. This is also envisaged in the context of the present invention.

[0248] It is further envisaged in the context of the present invention that the antigen-binding molecule is present in a concentration range of 0.1 to 8 mg / ml, preferably 0.2 to 2.5 mg / ml, more preferably 0.25 to 1.0 mg / ml.

[0249] It will be apparent to one skilled in the art that different components of a pharmaceutical composition (e.g., those listed above) may have different effects, e.g., amino acids may act as buffers, stabilizers, and / or antioxidants; mannitol may act as a bulking agent and / or tonicity enhancer; sodium chloride may act as a delivery vehicle and / or tonicity enhancer, etc.

[0250] It is envisaged that the composition of the present invention may contain, in addition to the polypeptide of the present invention as defined herein, further biologically active agents depending on the intended use of the composition. Such agents may be drugs acting on the gastrointestinal system, drugs acting as cytostatic agents, drugs preventing hyperuricemia, drugs inhibiting immune responses (e.g., corticosteroids), drugs modulating inflammatory responses, drugs acting on the circulatory system, and / or cytokines, etc., known in the art. It is also envisaged that the antigen-binding molecules of the present invention are applied in combination therapy, i.e. in combination with another anticancer drug.

[0251] In certain embodiments, the optimal pharmaceutical composition will be determined by those skilled in the art depending, for example, on the intended route of administration, delivery format, and desired dosage. See, for example, REMINGTON'S PHARMACEUTICAL SCIENCES, supra. In certain embodiments, such compositions may affect the physical state, stability, in vivo release rate, and in vivo clearance rate of the antigen binding protein of the present invention. In certain embodiments, the primary vehicle or carrier in the pharmaceutical composition may be either aqueous or non-aqueous in nature. For example, a suitable vehicle or carrier may be water for injection, saline solution, or artificial cerebrospinal fluid, optionally supplemented with other materials common in compositions for parenteral administration. Neutral buffered saline, or saline mixed with serum albumin, are further exemplary vehicles. In certain embodiments, the antigen-binding molecules of the compositions of the present invention can be prepared for storage by mixing the selected composition having the desired degree of purity in the form of a lyophilized cake or aqueous solution with optional compounding agents (REMINGTON'S PHARMACEUTICAL SCIENCES, supra). Furthermore, in certain embodiments, the antigen-binding molecules of the present invention can be formulated as a lyophilizate using appropriate excipients such as sucrose.

[0252] When parenteral administration is intended, the therapeutic composition for use in the present invention may be provided in the form of a pyrogen-free parenterally acceptable aqueous solution containing the desired antigen-binding molecule of the present invention in a pharma- ceutically acceptable vehicle. A particularly suitable vehicle for parenteral injection is sterile distilled water, in which the antigen-binding molecule of the present invention is formulated as a sterile isotonic solution appropriately preserved. In certain embodiments, the formulation may include a formulation of the desired molecule with an agent that can provide controlled or sustained release of a product that can be delivered via depot injection, such as injectable microspheres, biodegradable particles, polymeric compounds (e.g., polylactic acid or polyglycolic acid), beads, or liposomes. In certain embodiments, hyaluronic acid may also be used, which has the effect of enhancing duration in the circulation. In certain embodiments, an implantable drug delivery device may be used to introduce the desired antigen-binding molecule.

[0253] Further pharmaceutical compositions will be apparent to those skilled in the art, including formulations comprising the antigen-binding molecules of the present invention in sustained or controlled delivery / release formulations. Techniques for formulating various other sustained or controlled delivery means, such as liposome carriers, biodegradable microparticles or porous beads, and depot injections, are also known to those skilled in the art. See, for example, International Patent Application No. PCT / US93 / 00829, which describes the controlled release of porous polymeric microparticles for the delivery of pharmaceutical compositions. Sustained release formulations may include semipermeable polymer matrices in the form of shaped articles, such as films or microcapsules. The sustained release matrix may comprise polyesters, hydrogels, polylactides (disclosed in U.S. Pat. No. 3,773,919, EP 058481), copolymers of L-glutamic acid and gamma-ethyl-L-glutamate (Sidman et al., 1983, Biopolymers 2:547-556), poly(2-hydroxyethyl-methacrylate) (Langer et al., 1981, J. Biomed. Mater. Res. 15:167-277, and Langer, 1982, Chem. Tech. 12:98-105), ethylene vinyl acetate (Langer et al., 1981, supra), or poly-D(-)-3-hydroxybutyric acid (EP 133,988). The sustained release composition may also comprise liposomes, which can be prepared by any of several methods known in the art. See, e.g., Eppstein et al., 1985, Proc. Natl. Acad. Sci. USA 82:3688-3692; European Patent Applications 036,676; 088,046, and 143,949.

[0254] The antigen-binding molecules may also be encapsulated in microcapsules (e.g., hydroxymethylcellulose or gelatin-microcapsules, and poly(methyl methacrylate) microcapsules, respectively), colloidal drug delivery systems (e.g., liposomes, albumin microspheres, microemulsions, nanoparticles, and nanocapsules), or macroemulsions prepared, for example, by coacervation techniques or by interfacial polymerization. Such techniques are disclosed in Remington's Pharmaceutical Sciences, 16th edition, Oslo, A. Ed. (1980).

[0255] The pharmaceutical composition used for in vivo administration is typically provided as a sterile preparation. Sterilization can be achieved by filtration through a sterile filtration membrane. If the composition is lyophilized, sterilization using this method can be carried out either before or after lyophilization and reconstitution. The composition for parenteral administration can be stored in lyophilized form or in solution. Parenteral compositions are generally placed in a container with a sterile access port, for example, an intravenous solution bag or vial with a stopper that can be pierced by a hypodermic injection needle.

[0256] Another aspect of the present invention includes the self-buffering antigen-binding molecules of the formulations of the present invention that can be used as pharmaceutical compositions as described in International Patent Application WO 06138181 A2 (PCT / US2006 / 022599). Various descriptions are available on protein stabilization and formulation materials and methods useful in this regard, e.g., Arakawa et al., "Solvent interactions in pharmaceutical formulations," Pharm Res. 8(3):285-91 (1991); Kendrick et al., "Physical stabilization of proteins in aqueous solution" in: RATIONAL DESIGN OF STABLE PROTEIN FORMULATIONS: THEORY AND PRACTICE, Carpenter and Manning, eds. Pharmaceutical Biotechnology. 13:61-84 (2002), and Randolph et al., "Surfactant-protein interactions", Pharm Biotechnol. 13:159-75 (2002), particularly with respect to protein pharmaceuticals and processes for veterinary and / or human medical use, see in particular the sections relating to the same excipients and processes as the self-buffering protein formulations according to the present invention.

[0257] Salts may be used according to certain embodiments of the invention, for example, to adjust the ionic strength and / or tonicity of the formulation and / or to improve the solubility and / or physical stability of proteins or other components of the composition according to the invention. As is well known, ions may stabilize proteins in the native state by binding to charged residues on the surface of the protein and by shielding charged and polar groups in the protein, reducing the strength of their electrostatic, attractive and repulsive interactions. Ions may also stabilize proteins in the denatured state, in particular by binding to modified peptide bonds (--CONH) of the protein. Furthermore, ionic interactions with charged and polar groups in proteins may reduce intermolecular electrostatic interactions, thereby preventing or reducing protein aggregation and insolubilization.

[0258] Ionic species vary significantly in their effects on proteins. Several taxonomic rankings of ions and their effects on proteins have been developed that may be used in formulating pharmaceutical compositions according to the invention. One example is the Hofmeister series, which ranks ionic and polar non-ionic solutes by their effect on the conformational stability of proteins in solution. Stabilizing solutes are called "kosmotropics". Destabilizing solutes are called "chaotropics". Cosmotropes are commonly used at high concentrations (e.g., >1 molar ammonium sulfate) to precipitate ("salt out") proteins from solution. Chaotropes are commonly used to denature and / or solubilize ("salt in") proteins. The relative effects of an ion on "salting in" and "salting out" define the ion's position in the Hofmeister series.

[0259] Free amino acids may be used in the antigen-binding molecules of the formulations of the present invention according to various embodiments of the present invention as bulking agents, stabilizers, and antioxidants, and for other standard uses. Lysine, proline, serine, and alanine may be used to stabilize proteins in the formulation. Glycine is useful for ensuring proper cake structure and properties in lyophilization. Arginine may be useful for inhibiting protein aggregation in both liquid and lyophilized formulations. Methionine is useful as an antioxidant.

[0260] Polyols include sugars, such as mannitol, sucrose, and sorbitol, and polyhydric alcohols, such as glycerol and propylene glycol, and for purposes of discussion herein, polyethylene glycol (PEG) and related substances. Polyols are kosmotropic. They are useful stabilizers for protecting proteins from physical and chemical degradation processes in both liquid and lyophilized formulations. Polyols are also useful for adjusting the isotonicity of the formulation. Among the polyols useful in selected embodiments of the present invention is mannitol, which is commonly used in lyophilized formulations to ensure cake structural stability. Mannitol ensures cake structural stability. Generally, it is used in conjunction with a lyoprotectant, such as sucrose. Sorbitol and sucrose are among the preferred agents as stabilizers for adjusting isotonicity and for protecting against freeze-thaw stress during shipping or bulk preparation in the manufacturing process. Reducing sugars (containing free aldehyde or ketone groups), such as glucose and lactose, can glycate surface lysine and arginine residues. Therefore, they are generally not included among the preferred polyols for use in accordance with the present invention.In addition, sugars that form such reactive species, such as sucrose, are also not included among the preferred polyols of the present invention, in that they are hydrolyzed to fructose and glucose under acidic conditions, resulting in glycation.PEG is useful for stabilizing proteins and as a cryoprotectant, and in this regard may be used in the present invention.

[0261] The antigen-binding molecule embodiment of the formulation of the present invention further comprises a surfactant. Protein molecules may be prone to surface adsorption and denaturation and resulting aggregation at air-liquid, solid-liquid, and liquid-liquid interfaces. These effects are generally inversely proportional to protein concentration. These adverse interactions are generally inversely proportional to protein concentration and are typically exacerbated by physical agitation, such as that occurring during product transportation and handling. Surfactants are routinely used to prevent, minimize, or reduce surface adsorption. Surfactants useful in the present invention in this regard include polysorbate 20, polysorbate 80, other fatty acid esters of sorbitan polyethoxylate, and poloxamer 188. Surfactants are also commonly used to control protein conformational stability. Any given surfactant will usually stabilize some proteins and destabilize others, so the use of surfactants in this regard is protein specific.

[0262] Polysorbates are prone to oxidative degradation and often contain sufficient peroxide content when supplied to cause oxidation of the side chains of protein residues, particularly methionine. Polysorbates should therefore be used with caution and, when used, at the lowest possible concentrations. In this respect, polysorbates illustrate the general rule that excipients should be used at the lowest possible concentrations.

[0263] The antigen-binding molecule embodiment of the formulation of the present invention further comprises one or more antioxidants. By maintaining an appropriate level of ambient oxygen and temperature, and by avoiding exposure to light, harmful oxidation of proteins in pharmaceutical formulations can be prevented to some extent. Antioxidant excipients can also be used to prevent oxidative degradation of proteins. Particularly useful antioxidants in this regard include reducing agents, oxygen / free radical scavengers, and chelating agents. Antioxidants for use in therapeutic protein formulations according to the present invention are preferably water-soluble and maintain activity over the shelf life of the product. In this regard, EDTA is a preferred antioxidant according to the present invention. Antioxidants can damage proteins. For example, reducing agents (e.g., glutathione) can, inter alia, break intramolecular disulfide bonds. Thus, antioxidants for use in the present invention are selected, inter alia, to eliminate or sufficiently reduce the possibility of damaging proteins in the formulation.

[0264] The formulations according to the invention may contain metal ions that are cofactors for proteins and are required to form protein coordination compounds, for example zinc, which is required to form certain insulin suspensions. Metal ions may also inhibit some processes that break down proteins. However, metal ions also catalyze physical and chemical processes that break down proteins. Magnesium ions (10-120 mM) may be used to inhibit the isomerization of aspartic acid to isoaspartic acid. Ca +2 ions (up to 100 mM) can increase the stability of human deoxyribonuclease. +2 , Mn +2 , and Zn +2 can destabilize rhDNase. +2 and Sr +2 can stabilize factor VIII, which is +2 , Mn +2 and Zn +2 , Cu +2 and Fe +2 The aggregation can be destabilized by Al+3 It may be increased by ions.

[0265] The antigen-binding molecule embodiment of the formulation of the present invention further comprises one or more preservatives. Preservatives are necessary when developing a multi-dose parenteral formulation with more than one withdrawal from the same container. Its main function is to inhibit microbial growth and ensure sterility of the product over the shelf life or use period of the formulation. Commonly used preservatives include benzyl alcohol, phenol, and m-cresol. Although preservatives have a long history of use with small molecule parenteral drugs, developing protein formulations containing preservatives can be difficult. Preservatives almost always have a destabilizing effect on proteins (aggregation), which is the main factor limiting their use in multi-dose protein formulations. To date, most protein drugs have only been formulated for single use. However, if a multi-dose formulation were possible, it would have the added benefit of providing patient convenience and increased marketability. A good example is human growth hormone (hGH), where the development of a preserved formulation led to the commercialization of a more convenient multi-use injection pen proposal. At least four such pen devices containing preserved formulations of hGH are currently available on the market: Norditropin (liquid, Novo Nordisk), Nutropin AQ (liquid, Genentech) and Genotropin (lyophilized-dual chamber cartridge, Pharmacia & Upjohn) contain phenol, while Somatrope (Eli Lilly) is formulated with m-cresol. During the formulation and development of preserved dosage forms, several aspects must be considered: The effective preservative concentration in the drug product must be optimized. This requires testing a given preservative in the dosage form for a concentration range that confers antimicrobial efficacy without compromising protein stability.

[0266] As expected, developing liquid formulations containing preservatives is more challenging than lyophilized formulations. Lyophilized products can be lyophilized without preservatives and reconstituted with a diluent containing the preservative at the time of use. This reduces the time the preservative is in contact with the protein and greatly minimizes the associated stability risks. For liquid formulations, the efficacy and stability of the preservative should be maintained throughout the entire product shelf life (approximately 18-24 months). It is important to note that the efficacy of the preservative should be demonstrated in the final formulation containing the active drug and all excipient components.

[0267] The antigen-binding molecules disclosed herein may also be formulated as immunoliposomes. "Liposomes" are small vesicles composed of various types of lipids, phospholipids, and / or surfactants that are useful for drug delivery to mammals. The components of the liposome are generally arranged in a bilayer structure similar to the lipid arrangement of biological membranes. Liposomes containing antigen-binding molecules are prepared by methods known in the art, for example, as described in Epstein et al., Proc. Natl. Acad. Sci. USA, 82:3688 (1985); Hwang et al., Proc. Natl Acad. Sci. USA, 77:4030 (1980); U.S. Pat. Nos. 4,485,045 and 4,544,545; and WO 97 / 38731. Liposomes with enhanced circulation time are disclosed in U.S. Pat. No. 5,013,556. Particularly useful liposomes can be made by reverse phase evaporation using a lipid composition comprising phosphatidylcholine, cholesterol, and PEG-derivatized phosphatidylethanolamine (PEG-PE). Liposomes are extruded through filters of defined pore size to obtain liposomes with desired diameters. Fab' fragments of the antigen-binding molecules of the present invention can be conjugated to liposomes as described in Martin et al. J. Biol. Chem. 257:286-288 (1982) by disulfide exchange reaction. Optionally, chemotherapeutic agents are contained within the liposomes. See Gabizon et al. J. National Cancer Inst. 81(19)1484 (1989).

[0268] After the pharmaceutical composition has been formulated, it may be stored in sterile vials as a solution, suspension, gel, emulsion, solid, crystal, or as a dehydrated or lyophilized powder. Such formulations may be stored in a ready-to-use form or in a form (e.g., lyophilized) that is reconstituted prior to administration.

[0269] The biological activity of the pharmaceutical compositions defined herein may be determined, for example, by cytotoxicity assays as described in the Examples below, in WO 99 / 54440, or in Schlereth et al. (Cancer Immunol. Immunother. 20 (2005), 1-12). As used herein, "efficacy" or "in vivo efficacy" refers to the response to therapy with the pharmaceutical composition of the present invention, for example, using standardized NCI response criteria. The success or in vivo efficacy of therapy with the pharmaceutical composition of the present invention refers to the efficacy of the composition for its intended purpose, i.e., the ability of the composition to cause its desired effect, i.e., depletion of pathological cells (e.g., tumor cells). In vivo efficacy may be monitored by established standard methods for the respective disease entity, including, but not limited to, white blood cell counts, differential, fluorescence-activated cell sorting, bone marrow aspiration. In addition, various disease-specific clinical chemistry parameters and other established standard methods may be used.In addition, computed tomography, x-ray, and nuclear magnetic resonance imaging (e.g., response assessment based on the National Cancer Institute criteria [Cheson BD, Horning SJ, Coiffier B, Shipp MA, Fisher RI, Connors JM, Lister TA, Vose J, Grillo-Lopez A, Hagenbeek A, Cabanillas F, Klippensten D, Hiddemann W, Castellino R, Harris NL, Armitage JO, Carter W, Hoppe R, Canellos GP. Report of an international workshop to standardize response criteria for non-Hodgkin's lymphomas. NCI Sponsored International Working Group. J Clin Oncol. 1999 Apr;17(4):1244]), positron emission tomography scanning, white blood cell count, differential, fluorescence activated cell sorting, bone marrow aspiration, lymph node biopsy / histology, and various lymphoma-specific clinical chemistry parameters (e.g., lactate dehydrogenase), as well as other established standard methods may be used.

[0270] Another major challenge in the development of drugs, such as the pharmaceutical compositions of the present invention, is the predictable modulation of pharmacokinetic properties. To this end, a pharmacokinetic profile of a candidate drug can be established, i.e., a profile of pharmacokinetic parameters that affect the ability of a particular drug to treat a given pathology. Pharmacokinetic parameters of a drug that affect the ability of a drug to treat a certain disease include, but are not limited to, half-life, volume of distribution, hepatic first-pass metabolism, and degree of serum binding. The efficacy of a given drug can be influenced by each of the above-mentioned parameters. A presumed feature of the antigen-binding molecules of the present invention provided by a particular FC format is that they include differences, for example, in pharmacokinetic behavior. The targeted antigen-binding molecules with extended half-life of the present invention preferably exhibit a surprisingly increased residence time in vivo compared to the "canonical" non-HLE version of said antigen-binding molecule.

[0271] "Half-life" refers to the time it takes for 50% of an administered drug to be eliminated through biological processes, e.g., metabolism, excretion, etc. "Hepatic first-pass metabolism" refers to the tendency of a drug to be metabolized upon first contact with the liver, i.e., during its first passage through the liver. "Volume of distribution" refers to the degree of retention of a drug across various compartments of the body, e.g., intracellular and extracellular spaces, tissues and organs, etc., and the distribution of the drug within these compartments. "Extent of serum binding" refers to the tendency of a drug to interact with and bind to serum proteins, such as albumin, resulting in a reduction or elimination of the drug's biological activity.

[0272] Pharmacokinetic parameters also include bioavailability, lag time (Tlag), Tmax, absorption rate, onset of more action, and / or Cmax for a given amount of drug administered. "Bioavailability" refers to the amount of drug in the blood compartment. "Lag time" refers to the delay time from administration of a drug until the drug can be detected and measured in blood or plasma. "Tmax" is the time after which the maximum blood concentration of the drug is reached, and "Cmax" is the maximum blood concentration achieved by a given drug. All parameters affect the time it takes for a drug to reach the blood or tissue concentration required for a biological effect. Pharmacokinetic parameters of bispecific antigen-binding molecules exhibiting cross-species specificity that can be determined in preclinical animal studies in non-chimpanzee primates as outlined above are also described, for example, in Schlereth et al. (Cancer Immunol. Immunother. 20 (2005), 1-12).

[0273] In a preferred embodiment of the invention, the pharmaceutical composition is stable for at least 4 weeks at about -20°C. As is evident from the accompanying examples, the quality of the antigen-binding molecules of the invention relative to the quality of the corresponding state-of-the-art antigen-binding molecules can be tested using different systems. These tests are understood to be in accordance with the "ICH Harmonised Tripartite Guideline: Stability Testing of Biotechnological / Biological Products Q5C and Specifications: Test procedures and Acceptance Criteria for Biotech Biotechnological / Biological Products Q6B" and are selected to provide a stability indicator profile that provides reliable detection of changes in the identity, purity and potency of the product. It is well accepted that the term purity is a relative term. Due to the effects of glycosylation, deamidation or other heterogeneity, the absolute purity of a biotechnological / biological product should typically be assessed by multiple methods and the purity values ​​obtained are method-dependent. For the purpose of stability testing, the tests of purity should be aligned with the method of determination of degradation products.

[0274] To assess the quality of a pharmaceutical composition comprising an antigen-binding molecule of the present invention, it can be analyzed, for example, by analyzing the content of soluble aggregates in the solution (HMWS by size exclusion). Stability for at least 4 weeks at about -20°C is preferably characterized by a content of less than 1.5% HMWS, preferably less than 1% HMWS.

[0275] A preferred formulation for the antigen-binding molecule as a pharmaceutical composition may, for example, include the components of the formulation as follows: · formulation: Potassium phosphate pH 6.0, L-arginine hydrochloride, trehalose dihydrate, polysorbate 80.

[0276] Other examples of evaluation of the stability of the antigen-binding molecules of the present invention in the form of pharmaceutical compositions are provided in the accompanying Examples 4-12. In these examples, embodiments of the antigen-binding molecules of the present invention are tested for different stress conditions in different pharmaceutical formulations, and the results are compared with other half-life extension (HLE) format bispecific T cell engaging antigen-binding molecules known in the art. In general, it is assumed that the antigen-binding molecules provided in the specific FC format according to the present invention are typically more stable against a wide range of stress conditions, such as temperature and light stress, compared with both antigen-binding molecules provided in different HLE formats and antigen-binding molecules that do not have any HLE format (e.g., "canonical" antigen-binding molecules). The aforementioned temperature stability may relate to both low temperatures (below room temperature, including freezing temperatures) and high temperatures (above room temperature, including temperatures up to or above body temperature). As the skilled person will recognize, such improved stability against stresses that are difficult to avoid in clinical practice makes the antigen-binding molecules safer, since fewer degradation products are generated in clinical practice. As a result, the aforementioned improved stability means improved safety.

[0277] One embodiment provides an antigen binding molecule of the invention, or an antigen binding molecule produced according to the process of the invention, for use in the prevention, treatment, or amelioration of cancer associated with CD20 and CD22 expression or CD20 and CD22 overexpression, such as prostate cancer.

[0278] The formulations described herein are useful as pharmaceutical compositions for treating, ameliorating and / or preventing the pathological medical conditions described herein in patients in need thereof. The term "treatment" refers to both therapeutic treatment and prophylactic or preventative measures. Treatment includes application or administration of the formulation to the body, isolated tissues or cells of a patient having a disease / disorder, a symptom of a disease / disorder, or a predisposition to a disease / disorder, with the intent to cure, remedy, relieve, alleviate, alter, remedy, ameliorate, reverse or affect the disease, symptom of a disease, or predisposition to a disease.

[0279] As used herein, the term "amelioration" refers to any improvement in the disease state of a patient having a disease as described herein below, by administration of an antigen-binding molecule according to the present invention to a subject in need thereof. Such improvement may be seen as a slowing or halting of progression of the patient's disease. As used herein, the term "prevention" refers to the avoidance of the onset or recurrence of a tumor or cancer, or metastatic cancer, as described herein below, in a patient having such a tumor or cancer, by administration of an antigen-binding molecule according to the present invention to a subject in need thereof.

[0280] The term "disease" refers to any condition that would benefit from treatment with an antigen-binding molecule or pharmaceutical composition described herein, including chronic and acute disorders or diseases, including pathological conditions that predispose a mammal to the disease in question.

[0281] A "neoplasm" is an abnormal growth of tissue, usually, but not necessarily, forming a mass. When it forms a mass, it is also commonly referred to as a "tumor". A neoplasm or tumor can be benign, potentially malignant (precancerous), or malignant. A malignant neoplasm is generally referred to as a cancer. It usually invades and destroys surrounding tissues and can form metastases, i.e., it spreads to other parts, tissues, or organs of the body. Thus, the term "metastatic cancer" encompasses metastases to other tissues or organs other than that of the primary tumor. Lymphomas and leukemias are lymphatic neoplasms. For the purposes of the present invention, they are also encompassed by the terms "tumor" or "cancer".

[0282] The term "viral disease" refers to a disease that is the result of infection of a subject with a virus.

[0283] As used herein, the term "immune disorder" refers to immune disorders such as autoimmune diseases, hypersensitivity disorders, immune deficiencies, etc., in accordance with the general definition of this term.

[0284] In one embodiment, the present invention provides a method for the treatment or amelioration of cancer associated with expression of CD20 and CD22 or overexpression of CD20 and CD22, comprising administering to a subject in need thereof an antigen binding molecule of the present invention or an antigen binding molecule produced according to the method of the present invention. The CD20 and CD22xCD3 bispecific single chain antibodies are particularly advantageous for the treatment of cancer, preferably solid tumors, more preferably carcinomas and prostate cancer.

[0285] The term "subject in need" or "subject in need of treatment" includes subjects already with the disorder as well as those in which the disorder is to be prevented. A subject in need or "patient" includes human and other mammalian subjects receiving either prophylactic or therapeutic treatment.

[0286] The antigen-binding molecules of the present invention will generally be designed for a particular route and method of administration, a particular dosage and frequency of administration, and a particular treatment of a particular disease, particularly in the range of bioavailability and duration. The materials of the composition are preferably formulated at a concentration that is acceptable to the site of administration.

[0287] Thus, formulations and compositions may be designed for delivery by any suitable route of administration in accordance with the present invention. In the context of the present invention, routes of administration include, but are not limited to: · Topical route (e.g. on the skin, inhalation, nose, eyes, pinna / ear, vagina, mucous membranes); Enteral routes (e.g., oral, gastrointestinal, sublingual, sublabial, buccal, rectal); and Parenteral routes (e.g., intravenous, intraarterial, intraosseous, intramuscular, intracerebral, intraventricular, epidural, intrathecal, subcutaneous, intraperitoneal, extra-amniotic, intra-articular, intracardiac, intradermal, intralesional, intrauterine, intravesical, intravitreal, transdermal, intranasal, transmucosal, intrasynovial, intraluminal).

[0288] The pharmaceutical compositions and antigen-binding molecules of the present invention are particularly useful for parenteral administration, e.g., subcutaneous or intravenous delivery, e.g., by injection, e.g., bolus injection, or by infusion, e.g., continuous infusion. The pharmaceutical composition may be administered using a medical device. Examples of medical devices for administering pharmaceutical compositions are described in U.S. Pat. Nos. 4,475,196; 4,439,196; 4,447,224; 4,447,233; 4,486,194; 4,487,603; 4,596,556; 4,790,824; 4,941,880; 5,064,413; 5,312,335; 5,312,335; 5,383,851; and 5,399,163.

[0289] In particular, the present invention provides for uninterrupted administration of suitable compositions. As a non-limiting example, uninterrupted or substantially uninterrupted, i.e. continuous administration, can be achieved by a miniature pump system worn by the patient to regulate the inflow of therapeutic agent into the patient's body. The pharmaceutical composition comprising the antigen-binding molecule of the present invention can be administered by using said pump system. Such pump systems are generally known in the art and typically rely on periodic replacement of a cartridge containing the therapeutic agent to be infused. When replacing the cartridge in such a pump system, a temporary interruption in the otherwise uninterrupted inflow of therapeutic agent into the patient's body may result. Even in such a case, the administration step before the cartridge replacement and the administration step after the cartridge replacement will still be considered within the meaning of the pharmaceutical means and methods of the present invention, both of which constitute the "uninterrupted administration" of such therapeutic agent.

[0290] The continuous or uninterrupted administration of the antigen-binding molecules of the present invention may be intravenous or subcutaneous administration by a fluid delivery device or a miniature pump system that includes a fluid delivery mechanism for pumping fluid from a reservoir and a drive mechanism for driving the delivery mechanism. The pump system for subcutaneous administration may include a needle or cannula for penetrating the patient's skin and delivering the suitable composition into the patient's body. The pump system may be fixed or attached directly to the patient's skin, whether vein, artery, or blood vessel, allowing direct contact between the pump system and the patient's skin. The pump system may be attached to the patient's skin for 24 hours to several days. In some cases, the pump system may be a miniature pump system with a small reservoir volume. As a non-limiting example, the reservoir volume for the suitable pharmaceutical composition to be administered may be 0.1 to 50 ml.

[0291] Continuous administration can also be transdermal, by a patch that is attached to the skin and replaced from time to time.Those skilled in the art are aware of suitable patch systems for drug delivery for this purpose.It should be noted that transdermal administration is particularly suitable for uninterrupted administration, since it has the advantage that, for example, a new second patch can be attached to the skin surface immediately adjacent to the first used patch, and the replacement of the first used patch can be completed at the same time just before removing the first used patch.The problem of interruption of inflow or battery failure does not occur.

[0292] If the pharmaceutical composition is lyophilized, the lyophilized material is first reconstituted with a suitable liquid prior to administration, which may be, for example, bacteriostatic water for injection (BWFI), saline, phosphate buffered saline (PBS), or the same formulation in which the protein was present prior to lyophilization.

[0293] The composition of the present invention may be administered to a subject at a suitable dose, which may be determined, for example, by a dose escalation study in which the antigen-binding molecules of the present invention exhibiting cross-species specificity as described herein are administered to non-chimpanzee primates, such as macaques, in increasing doses. As described above, the antigen-binding molecules of the present invention exhibiting cross-species specificity as described herein have the advantage that they can be used in the same form in preclinical trials in non-chimpanzee primates and as drugs in humans. The administration schedule will be determined by the attending physician based on clinical factors. As is known in the medical field, the dosage for a given patient depends on many factors, including the patient's size, body surface area, age, the individual compound administered, sex, time and route of administration, general health, and other drugs administered at the same time.

[0294] The term "effective dose" or "effective dosage" is defined as an amount sufficient to achieve or at least partially achieve a desired effect. The term "therapeutically effective dose" is defined as an amount sufficient to cure or at least partially arrest the disease and its complications in a patient already suffering from the disease. The amount or dosage effective for this use will depend on the condition (indication) being treated, the antigen-binding molecule being delivered, the nature and purpose of the treatment, the severity of the disease, previous treatments, the patient's medical history and response to the therapeutic agent, the route of administration, the size (weight, body surface area, or organ size), and / or the condition (age and general health) of the patient, and the general condition of the patient's own immune system. The appropriate dose may be administered to the patient in a single dose or multiple doses, or may be adjusted according to the judgment of the attending physician to obtain the optimal therapeutic effect.

[0295] Typical dosages can range from about 0.1 μg / kg up to about 30 mg / kg or more, depending on the factors mentioned above. In certain embodiments, dosages can range from 1.0 μg / kg up to about 20 mg / kg, optionally 10 μg / kg up to about 10 mg / kg, or 100 μg / kg up to about 5 mg / kg.

[0296] A therapeutically effective amount of the antigen binding molecule of the present invention preferably reduces the severity of disease symptoms, increases the frequency or duration of disease symptom-free periods, or prevents functional or disability impairments due to disease suffering. For treating diseases correlated with CD20 and CD22 expression as described herein above, a therapeutically effective amount of the antigen binding molecule of the present invention, here anti-CD20 and CD22 / anti-CD3 antigen binding molecule, preferably inhibits cell growth or tumor growth by at least about 20%, at least about 40%, at least about 50%, at least about 60%, at least about 70%, at least about 80%, or at least about 90% compared to untreated patients. The ability of the compound to inhibit tumor growth can be evaluated in animal models predicting efficacy.

[0297] The pharmaceutical composition may be administered as a single treatment or, if necessary, in combination with additional therapies, such as anti-cancer therapies, e.g., other proteinaceous and non-proteinaceous drugs. These drugs may be administered simultaneously with the composition comprising the antigen-binding molecule of the present invention as defined herein, or may be administered separately at defined time intervals and doses before or after administration of said antigen-binding molecule.

[0298] As used herein, the term "effective and non-toxic dose" refers to a tolerable dose of the antigen-binding molecule of the present invention that is high enough to cause depletion of pathological cells, tumor disappearance, tumor regression, or disease stabilization without or essentially without causing significant toxic effects. Such an effective and non-toxic dose may be determined, for example, by dose escalation studies as described in the art, and should be below the dose that induces serious adverse side events (dose-limiting toxicity, DLT).

[0299] As used herein, the term "toxicity" refers to the toxic effects of a drug that manifest as an adverse event or a serious adverse event. This side event may refer to a lack of systemic drug tolerance and / or a lack of local tolerance after administration. Toxicity may also include teratogenic or carcinogenic effects caused by the drug.

[0300] As used herein, the term "safety", "in vivo safety", or "tolerability" defines the administration of a drug that does not induce serious adverse events immediately after administration (local tolerance) and during longer drug application periods. "Safety", "in vivo safety", or "tolerability" may be evaluated, for example, during treatment and periodically during follow-up. Measurements include clinical evaluations, for example, organ findings, and screening for laboratory abnormalities. Clinical evaluations may be performed and deviations from normal findings may be recorded / coded according to NCI-CTC and / or MedDRA standards. Organ findings may include criteria such as allergy / immunology, blood / bone marrow, cardiac arrhythmias, coagulation, etc., as set forth in, for example, the Common Terminology Criteria for adverse events v3.0 (CTCAE). Laboratory parameters that may be tested include, for example, hematology, clinical chemistry, coagulation profile, and urinalysis, as well as tests of other body fluids (e.g., serum, plasma, lymphatic or spinal fluid, cerebrospinal fluid, etc.). Thus, safety may be assessed, for example, by physical examination, imaging techniques (i.e. ultrasound, x-ray, CT scan, magnetic resonance imaging (MRI), other measurements using technical devices (i.e. electrocardiogram), vital signs, measuring laboratory parameters and recording adverse events. For example, in the uses and methods according to the invention, adverse events in non-chimpanzee primates may be tested by histopathological and / or histochemical methods.

[0301] The above terms are also referenced, for example, in Preclinical safety evaluation of biotechnology-derived pharmaceuticals S6; ICH Harmonised Tripartite Guideline; ICH Steering Committee meeting of July 16, 1997.

[0302] Finally, the present invention provides a kit comprising an antigen-binding molecule of the present invention or produced according to the process of the present invention, a pharmaceutical composition of the present invention, a polynucleotide of the present invention, a vector of the present invention, and / or a host cell of the present invention.

[0303] In the context of the present invention, the term "kit" refers to two or more components packaged together in a container, vessel, or other package, one of which corresponds to an antigen-binding molecule, pharmaceutical composition, vector, or host cell of the present invention. Thus, a kit can be described as a set of products and / or tools sufficient to achieve a particular purpose that can be sold as a single item.

[0304] The kit may include one or more containers (e.g., vials, ampoules, containers, syringes, bottles, bags) of any suitable shape, size, and material (preferably waterproof, e.g., plastic or glass) containing the antigen-binding molecule or pharmaceutical composition of the present invention in a dosage amount suitable for administration (see above). The kit may further include instructions for use (e.g., in the form of a leaflet or instruction manual), a means for administering the antigen-binding molecule of the present invention, such as a syringe, pump, infuser, etc., a means for reconstituting the antigen-binding molecule of the present invention, and / or a means for diluting the antigen-binding molecule of the present invention.

[0305] The present invention also provides a kit for single-dose administration unit. The kit of the present invention may also include a first container containing a dried / lyophilized antigen-binding molecule and a second container containing an aqueous formulation. In certain embodiments of the present invention, a kit is provided that includes single-chamber and multi-chamber pre-filled syringes (e.g., liquid syringes and dissolution syringes).

[0306] It should be noted that, as used herein, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, a reference to a "reagent" includes one or more of such various reagents, and a reference to a "method" includes references to equivalent steps and methods known to those of skill in the art that may be modified for or substituted for the method described herein.

[0307] Unless otherwise indicated, the term "at least" preceding a series of elements should be understood to refer to every element in the series. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the present invention.

[0308] The term "and / or" wherever used in this specification includes the meaning of "and", "or" and "all or any other combination of the elements connected by said term".

[0309] As used herein, the term "about" or "approximately" means within 20%, preferably within 10%, and more preferably within 5% of a given value or range, however, the term also includes specific numbers, for example, about 20 includes 20.

[0310] The terms "less than" or "greater than" include a specific number. For example, less than 20 means less than or equal to. Similarly, greater than or greater than means greater than or equal to, or greater than or equal to, respectively.

[0311] Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise" and variations such as "comprises" and "comprising" should be understood to mean the inclusion of a stated integer or step or group of integers or steps, but not the exclusion of any other integer or step or group of integers or steps. As used herein, the term "comprising" may also be replaced with the terms "containing" or "including," or, as sometimes used herein, the term "having."

[0312] As used herein, "consisting of" excludes any element, step, or ingredient not specified in the claim element. As used herein, "consisting essentially of" does not exclude materials or steps that do not materially affect the basic and novel characteristics of the claim.

[0313] As used herein, in each instance, any of the terms "comprising," "consisting essentially of," and "consisting of" may be replaced with either of the other two terms.

[0314] It is to be understood that the invention is not limited to the particular methodology, protocols, materials, reagents, and substances, etc. described herein and as such may vary. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of the invention, which is defined solely by the claims.

[0315] All publications and patents cited throughout the text of this specification, whether supra or infra, including all patents, patent applications, scientific publications, manufacturer's specifications, instructions, etc., are hereby incorporated by reference in their entirety. Nothing herein should be construed as an admission that the invention is not entitled to antedate such disclosure by virtue of prior invention. To the extent that material incorporated by reference is inconsistent or inconsistent with the present specification, the present specification will take precedence over any such material.

[0316] A better understanding of the present invention and its advantages will be obtained from the following examples, which are provided for illustrative purposes only and are not intended to limit the scope of the invention in any way. EXAMPLES

[0317] Example 1: Evaluation of productivity and product uniformity Protein purification by two-step fast protein liquid chromatography An Aekta Pure Purification System (Cytiva Life Sciences) controlled by Unicorn® 7.3 software was used for affinity capture and size-exclusion chromatography according to the manufacturer's specifications.

[0318] Protein isolation by affinity capture (AC) chromatography Capture of C20 and CD22 targeted antigen binding molecules was performed using HiTrap MabSelect SuRe® (5 ml column volume (CV); Cytiva Life Sciences) Protein A affinity media. The column was equilibrated with 2 CV phosphate buffered saline (PBS; Ca2+ and Mg2+ free; EMD Millipore) and protein-containing cell culture supernatant was applied to the column at a flow rate of 6 ml / min. Prior to protein elution, the column was washed sequentially with PBS and 0.5 M L-arginine, 25 mM Tris, pH 7.5 (10 CV each) to remove unbound or weakly bound host cell proteins. Bound proteins were eluted by application of 3 CV Protein A IgG elution buffer (90 mM NaCl, 20 mM citric acid, pH 3.0) at a flow rate of 2 ml / min and 6 ml of eluate was collected in the attached sample loop.

[0319] Isolation of protein monomers by size exclusion chromatography (SEC) Following AC, the proteins were transferred from the sample loop to a HiLoad S200 26 / 600 Superdex Gelfiltration SEC column (320 ml CV; Cytiva Life sciences) pre-equilibrated with 1.5 CV of formulation buffer (10 mM citric acid, 75 ml lysine HCl, pH 7.0). Monomeric proteins were then separated from HMW and LMW protein species by applying 1.5 CV of formulation buffer at a flow rate of 2.5 ml / min, and finally collected in a fraction collector.

[0320] For protein stabilization, trehalose was added to each collected fraction containing monomer to a final concentration of 4% trehalose. In addition, protein concentration was determined using A280 nm light absorption, and fractions with sufficient enrichment of monomeric protein were pooled. After concentration to 0.25 mg / ml and filtration, the yield of pure monomeric protein was calculated based on the total protein amount. The SEC peak symmetry of the main monomeric peak was obtained by the software Unicorn® 7.3 software at half maximum peak height.

[0321] [Table 4]

[0322] Final protein monomer yield and SEC monomer peak symmetry of CD20 and CD22 targeting antigen binding molecules. Yields were calculated based on total protein amount after purification, filtration, and concentration to 0.25 mg / ml. SEC peak symmetry was calculated with Unicorn software.

[0323] result All selected CD20 and CD22 dual targeting antigen binding molecules according to the present invention show a productivity of more than 10 mg / L in terms of final yield, in contrast to the comparative molecule, CD20 99-E5 CC×CD22 28-B7 N65S CC×I2C0×scFc. The molecules according to the present invention also show a homogeneous structure compared to the comparative molecule CD20 99-E5 CC×CD22 28-B7 N65S CC×I2C0×scFc, with a dynamic radius lower than the preferred threshold of 1.4. The symmetric peak of these new molecules suggests less low molecular weight products or less folded forms, thus improving product homogeneity.

[0324] Example 2 Evaluation of the hydrophobicity of the surface of CD20-CD22-targeting antigen-binding molecules The isolated and formulated CD20-CD22 binding T cell engager molecules and monomers adjusted to defined protein concentrations were transferred into autosampler-fitted sample vials and measured on an Aekta Purifier 10 FPLC system (GE Healthcare, Freiburg, Germany). A hydrophobic interaction chromatography HIC column was equilibrated with formulation buffer and a defined volume of protein solution was applied at a constant formulation buffer flow rate. Detection was performed by optical absorption at OD280 nm. Elution behavior was determined by peak shape and the slope of the decaying signal peak was mathematically calculated, respectively. A steeper / higher slope value indicates less hydrophobic interactions on the protein surface compared to constructs with flatter elution behavior and lower slope values.

[0325] [Table 5]

[0326] As can be seen from Table 5, for molecules according to the invention, HEC elution gradients of more than 15 (typically more than 25) can be observed. The higher the gradient, the less hydrophobic the product and therefore the better the productivity and stability.

[0327] In vitro affinity evaluation of CD20-CD22 dual-targeting antigen-binding molecules The cell-based affinity of the CD20 CD22 dual targeting antigen binding molecules was determined by nonlinear regression (single-site specific binding) analysis. CHO cells expressing human CD20, cynomolgus CD20, human CD22, or cynomolgus CD22 were incubated with decreasing concentrations of CD20 CD22 dual targeting antigen binding molecules (up to 800 nM, step 1:2 or 1:3, 11 steps) for 16 hours at 4°C. Bound CD20 CD22 dual targeting antigen binding molecules were detected with Alexa Flour 488-conjugated AffiniPure Fab fragment goat anti-human IgG (H+L). Fixed cells were stained with DRAQ5, Far-Red fluorescent live cell permeant DNA dye, and signals were detected by fluorescence cytometry. The respective equilibrium dissociation constant (Kd) values ​​were calculated using the single-site specific binding evaluation tool of GraphPad Prism software. The average Kd values ​​and affinity gaps were calculated in Microsoft Excel.

[0328] [Table 6]

[0329] The cell-based affinity of the CD20 CD22 dual targeting antigen binding molecule to target transfected CHO cells was determined by nonlinear regression (single-site specific binding) analysis. The average Kd value was calculated from three independent measurements. The affinity gap was determined by dividing the cynomolgus monkey Kd by the human Kd.

[0330] result Measurement of cell-based affinity revealed that the CD20 CD22 dual-targeting antigen-binding molecules 2 to 16 had higher cell-based affinity for human or cynomolgus monkey CD20-positive CHO cells and a smaller cynomolgus monkey / human gap in CD22-positive CHO cells than the CD20 CD22 dual-targeting antigen-binding molecule 1.

[0331] FACS-based cytotoxicity assay with unstimulated human PBMCs Isolation of effector cells Human peripheral blood mononuclear cells (PBMCs) were prepared by Ficoll density gradient centrifugation from concentrated lymphocyte preparations (buffy coats), a by-product of blood banks that collect blood for transfusion. Buffy coats were provided by peripheral blood banks, and PBMCs were prepared on the same day of blood collection. After Ficoll density centrifugation and extensive washing with Dulbecco's PBS (Gibco), residual red blood cells were removed from PBMCs via incubation with red blood cell lysis buffer (155 mM NH4Cl, 10 mM KHCO3, 100 μM EDTA). Centrifugation of PBMCs at 100 × g removed platelets via the supernatant. The remaining lymphocytes included mainly B and T lymphocytes, NK cells, and monocytes. PBMCs were maintained in RPMI medium (Gibco) containing 10% FCS (Gibco) at 37 °C / 5% CO2 in culture.

[0332] Depletion of CD14+, CD15+, CD16+, CD19+, CD34+, CD36+, CD56+, CD123+, and CD235a+.

[0333] For depletion of CD14+, CD15+, CD16+, CD19+, CD34+, CD36+, CD56+, CD123+, and CD235a+ cells, Human Pan T Cell Isolation Kit (Miltenyi Biotec, #130-096-535) was used. PBMCs were counted and centrifuged at 300×g for 10 min at room temperature. The supernatant was discarded and the cell pellet was resuspended in MACS isolation buffer [80 μL / 107 cells; PBS (Invitrogen, #20012-043), 0.5% (v / v) FBS (Gibco, #10270-106), 2 mM EDTA (Sigma-Aldrich, #E-6511)]. Human Pan T Cell Isolation Kit (20 μL / 107 cells) was added and incubated for 15 min at 4-8°C. Cells were washed with MACS isolation buffer (1-2 ml / 107 cells). After centrifugation (see above), the supernatant was discarded and cells were resuspended in MACS isolation buffer (500 μL / 108 cells). CD14, CD15, CD16, CD19, CD34, CD36, CD56, CD123, and CD235a negative cells were then isolated using LS columns (Miltenyi Biotec, #130-042-401). Pan T cells were cultured in RPMI complete medium, i.e. RPMI1640 (Biochrom AG, #FG1215) supplemented with 10% FBS (Bio West, #S0115), 1× non-essential amino acids (Biochrom AG, #K0293), 10 mM Hepes buffer (Biochrom AG, #L1613), 1 mM sodium pyruvate (Biochrom AG, #L0473), and 100 U / mL penicillin / streptomycin (Biochrom AG, #A2213), at 37° C. in an incubator until needed.

[0334] Target cell labeling For analysis of cell lysis in flow cytometry assays, fluorescent membrane dye DiOC18 (DiO) (Molecular Probes, #V22886) was used to label human CD20 and CD22 double positive human cell line Oci-Ly 1, human CD20 single positive human cell line Oci-Ly 1 (CD22 knockout clone #A1), and CD22 single positive human cell line Oci-Ly 1 (CD20 knockout clone #A5) as target cells and to distinguish these cell lines from effector cells. Briefly, cells were harvested, washed once with PBS, and adjusted to 106 cells / mL with PBS containing 2% (v / v) FBS and membrane dye DiO (5 μL / 106 cells). After 3 min incubation at 37°C, cells were washed twice with complete RPMI medium and cell number was adjusted to 1.25×105 cells / mL. Cell vitality was determined using an NC-250 cell counter (Chemometec).

[0335] Flow cytometry-based analysis The assay was designed to quantify the lysis of Oci-Ly 1 cells in the presence of serial dilutions of CD20 and CD22 dual targeting antigen binding molecules. Equal volumes of DiO-labeled target cells and effector cells (i.e. panT cells) were mixed to obtain an E:T cell ratio of 10:1. 80 μl of this suspension was transferred to each well of a 96-well plate. 20 μL of serial dilutions of CD20 and CD22 dual targeting antigen binding molecules were added, as well as a negative control (CD3-based T cell engager molecule recognizing an unrelated target antigen) or RPMI complete medium as an additional negative control. The cytotoxic reaction of the dual targeting antigen binding molecules was allowed to proceed for 48 hours in a humidified incubator at 7% CO2. The cells were then transferred to a new 96-well plate and the loss of target cell membrane integrity was monitored by adding propidium iodide (PI) to a final concentration of 1 μg / mL. PI is a membrane-impermeable dye that is normally excluded from viable cells, whereas dead cells take it up and become identifiable by fluorescence emission.

[0336] Samples were measured by flow cytometry on an iQue Plus instrument and analyzed by Forecyt software (both Intellicyt). Target cells were identified as DiO positive cells. PI negative target cells were classified as viable target cells. The percentage of cytotoxicity was calculated by the following formula:

[0337]

number

[0338] The percentage of cytotoxicity was plotted against the corresponding CD20 and CD22 dual targeting antigen binding molecule concentration using GraphPad Prism 5 software (Graph Pad Software, San Diego). The dose-response curves were analyzed using a four-parametric logistic regression model for the evaluation of sigmoidal dose-response curves with a fixed Hill slope, and EC50 values ​​were calculated.

[0339] [Table 7]

[0340] Table 7 shows the 48-hour FACS-based cytotoxicity assay of CD20 and CD22 dual-targeting antigen binding molecules with human CD20 and CD22 double-positive human cell line Oci-Ly 1, human CD20 single-positive human cell line Oci-Ly 1 (CD22 knockout clone #A1), and CD22 single-positive human cell line Oci-Ly 1 (CD20 knockout clone #A5) as target cells, and panT (E:T ratio 10:1) as effector cells. EC50 values ​​are determined by a four-parametric logistic regression model for the evaluation of sigmoidal dose-response curves with a fixed Hill slope.

[0341] Cytotoxicity assays with human CD20 and CD22 double positive human Oci-Ly 1 cells revealed that all conjugates showed better biological activity in the 1-2 pM range compared to the conjugate CD20 99-E5 CC×CD22 28-B7 N65S CC×I2C0×scFc(G3P).

[0342] Table 8: Sequence Listing The following table shows the sequences of the whole antigen-binding molecule and its fragments and / or building blocks. In each sequence description, I2C stands for CD3 effector binding domain. I2E stands for CD3 effector binding domain with increased stability. HLE stands for half-life extension domain, typically a scFc domain. scFv stands for a combination of VH and VL that together form a functional target or effector binding domain. Bispecific molecules stand for a combination of at least one target binding domain and one effector binding domain that together form a functional bispecific antigen-binding molecule. Target is typically abbreviated by two letters.

[0343] [Table 8]

[0344] [Table 9]

[0345] [Table 10]

[0346] [Table 11]

[0347] [Table 12]

[0348]

Table 13

[0349]

Table 14

[0350]

Table 15

[0351]

Table 16

[0352]

Table 17

[0353]

Table 18

[0354]

Table 19

[0355]

Table 20

[0356]

Table 21

[0357]

Table 22

[0358]

Table 23

[0359]

Table 24

[0360]

Table 25

[0361]

Table 26

[0362]

Table 27

[0363]

Table 28

[0364]

Table 29

[0365]

Table 30

[0366]

Table 31

[0367]

Table 32

[0368]

Table 33

[0369]

Table 34

[0370]

Table 35

[0371]

Table 36

[0372]

Table 37

[0373]

Table 38

[0374]

Table 39

[0375]

Table 40

[0376]

Table 41

[0377]

Table 42

[0378]

Table 43

[0379]

Table 44

[0380]

Table 45

[0381]

Table 46

[0382]

Table 47

[0383]

Table 48

[0384]

Table 49

[0385]

Table 50

[0386]

Table 51

[0387]

Table 52

[0388]

Table 53

[0389]

Table 54

[0390]

Table 55

[0391]

Table 56

[0392]

Table 57

[0393]

Table 58

[0394]

Table 59

[0395]

Table 60

[0396]

Table 61

[0397]

Table 62

[0398]

Table 63

[0399]

Table 64

[0400]

Table 65

[0401]

Table 66

[0402]

Table 67

[0403]

Table 68

[0404]

Table 69

[0405]

Table 70

[0406]

Table 71

[0407]

Table 72

[0408]

Table 73

[0409]

Table 74

Claims

1. A CD20 and CD22 targeting antigen binding molecule comprising at least three binding domains, (i.) a first binding domain comprises a paratope that immunospecifically binds to CD20, said first binding domain comprising: a) CDRs H1-3 of SEQ ID NOs: 58-60 and CDRs L1-3 of SEQ ID NOs: 61-63; b) CDRs H1-3 of SEQ ID NOs: 71-73 and CDRs L1-3 of SEQ ID NOs: 74-76; c) CDRs H1-3 of SEQ ID NOs: 84-86 and CDRs L1-3 of SEQ ID NOs: 87-89, and d) CDRs H1-3 of SEQ ID NOs: 97-99 and CDRs L1-3 of SEQ ID NOs: 100-102 a VH region comprising CDR-H1, CDR-H2, and CDR-H3, and a VL region comprising CDR-L1, CDR-L2, and CDR-L3 selected from (ii.) a second binding domain comprises a paratope that immunospecifically binds to CD22, said second binding domain comprising: a) CDR H1-3 of SEQ ID NOs: 138-140 and CDR L1-3 of SEQ ID NOs: 141-143; b) CDRs H1-3 of SEQ ID NOs: 151-153 and CDRs L1-3 of SEQ ID NOs: 154-156; c) CDRs H1-3 of SEQ ID NOs: 164-166 and CDRs L1-3 of SEQ ID NOs: 167-169; d) CDRs H1-3 of SEQ ID NOs: 177-179 and CDRs L1-3 of SEQ ID NOs: 180-182; e) CDRs H1-3 of SEQ ID NOs: 190-192 and CDRs L1-3 of SEQ ID NOs: 193-195; f) CDR H1-3 of SEQ ID NOs: 203-205 and CDR L1-3 of SEQ ID NOs: 206-208; g) CDRs H1-3 of SEQ ID NOs: 125-127 and CDRs L1-3 of SEQ ID NOs: 128-130; h) CDRs H1-3 of SEQ ID NOs: 216-218 and CDRs L1-3 of SEQ ID NOs: 219-221, and i) CDR H1-3 of SEQ ID NOs: 379-381 and CDR L1-3 of SEQ ID NOs: 382-384 a VH region comprising CDR-H1, CDR-H2, and CDR-H3, and a VL region comprising CDR-L1, CDR-L2, and CDR-L3 selected from (iii.) the third binding domain comprises a paratope that immunospecifically binds to an extracellular epitope of the human and / or macaque CD3 epsilon chain; the first, second, and third binding domains are arranged in amino to carboxyl order, and the first binding domain and the second binding domain are linked by a peptide linker having a length of 5 to 24 amino acids; the antigen-binding molecule comprises a fourth domain, the fourth domain comprising two polypeptide monomers, each polypeptide monomer comprising a hinge, a CH2 domain, and a CH3 domain, the two polypeptide monomers being fused to each other via a peptide linker; The fourth domain is composed of, in order from amino to carboxyl: Hinge-CH2-CH3-linker-hinge-CH2-CH3 and / or each of the polypeptide monomers in the fourth domain has an amino acid sequence that is at least 90% identical to a sequence selected from the group consisting of SEQ ID NOs: 17-24; and / or the CH2 domain comprises an intradomain cysteine ​​disulfide bridge; and / or the first, second, third, and fourth binding domains are arranged in amino to carboxyl order; CD20 and CD22 targeting antigen binding molecules.

2. The CD20 and CD22 targeted antigen binding molecule described in claim 1, wherein each of the polypeptide monomers has an amino acid sequence selected from SEQ ID NOs: 17 to 24.

3. The CD20- and CD22-targeting antigen-binding molecule of claim 1 or 2, wherein the antigen-binding molecule is a single-chain CD20- and CD22-targeting scFv antigen-binding molecule.

4. 4. The CD20 and CD22 targeted antigen binding molecule of any one of claims 1 to 3, wherein the peptide linker between the first binding domain and the second binding domain is selected from those having a length of 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, or 24 amino acids, or 5, 6, 7, 8, 9, 10, 11, or 12 amino acids.

5. The peptide linker between the first binding domain and the second binding domain is S(G ４ S) ｎ , (G ４ S) ｎ , G ４ｎ , and G ５ｎ (wherein n is equal to 1, 2, 3, or 4, or n is equal to 1 or 2), or the peptide linker is selected from the group consisting of: SG ４ The CD20 and CD22 targeted antigen binding molecule of any one of claims 1 to 4, wherein said antigen binding molecule is S.

6. The first binding domain and the second binding domain each comprise: a) CDRs H1-3 of SEQ ID NOs: 58-60, and CDRs L1-3 of SEQ ID NOs: 61-63 of said first binding domain, and CDRs H1-3 of SEQ ID NOs: 138-140, and CDRs L1-3 of SEQ ID NOs: 141-143 of said second binding domain; b) CDRs H1-3 of SEQ ID NOs: 58-60 and CDRs L1-3 of SEQ ID NOs: 61-63 of said first binding domain and CDRs H1-3 of SEQ ID NOs: 151-153 and CDRs L1-3 of SEQ ID NOs: 154-156 of said second binding domain; c) CDRs H1-3 of SEQ ID NOs: 58-60, and CDRs L1-3 of SEQ ID NOs: 61-63 of the first binding domain, and CDRs H1-3 of SEQ ID NOs: 164-166, and CDRs L1-3 of SEQ ID NOs: 167-169 of the second binding domain; d) CDRs H1-3 of SEQ ID NOs: 58-60 and CDRs L1-3 of SEQ ID NOs: 61-63 of the first binding domain, and CDRs H1-3 of SEQ ID NOs: 177-179 and CDRs L1-3 of SEQ ID NOs: 180-182 of the second binding domain; e) CDRs H1-3 of SEQ ID NOs: 58-60, and CDRs L1-3 of SEQ ID NOs: 61-63 of the first binding domain, and CDRs H1-3 of SEQ ID NOs: 190-192, and CDRs L1-3 of SEQ ID NOs: 193-195 of the second binding domain; f) CDRs H1-3 of SEQ ID NOs: 58-60 and CDRs L1-3 of SEQ ID NOs: 61-63 of the first binding domain, and CDRs H1-3 of SEQ ID NOs: 203-205 and CDRs L1-3 of SEQ ID NOs: 206-208 of the second binding domain; g) CDRs H1-3 of SEQ ID NOs: 58-60 and CDRs L1-3 of SEQ ID NOs: 61-63 of the first binding domain and CDRs H1-3 of SEQ ID NOs: 125-127 and CDRs L1-3 of SEQ ID NOs: 128-130 of the second binding domain; h) CDRs H1-3 of SEQ ID NOs: 58-60 and CDRs L1-3 of SEQ ID NOs: 61-63 of the first binding domain, and CDRs H1-3 of SEQ ID NOs: 216-218 and CDRs L1-3 of SEQ ID NOs: 219-221 of the second binding domain; i) CDRs H1-3 of SEQ ID NOs: 71-73, and CDRs L1-3 of SEQ ID NOs: 74-76 of said first binding domain, and CDRs H1-3 of SEQ ID NOs: 379-381, and CDRs L1-3 of SEQ ID NOs: 382-384 of said second binding domain; j) CDRs H1-3 of SEQ ID NOs: 71-73, and CDRs L1-3 of SEQ ID NOs: 74-76 of the first binding domain, and CDRs H1-3 of SEQ ID NOs: 203-205, and CDRs L1-3 of SEQ ID NOs: 206-208 of the second binding domain; k) CDRs H1-3 of SEQ ID NOs: 84-86 and CDRs L1-3 of SEQ ID NOs: 87-89 of the first binding domain and CDRs H1-3 of SEQ ID NOs: 164-166 and CDRs L1-3 of SEQ ID NOs: 167-169 of the second binding domain; l) CDRs H1-3 of SEQ ID NOs: 97-99 and CDRs L1-3 of SEQ ID NOs: 100-102 of the first binding domain, and CDRs H1-3 of SEQ ID NOs: 177-179 and CDRs L1-3 of SEQ ID NOs: 180-182 of the second binding domain; m) CDRs H1-3 of SEQ ID NOs: 97-99 and CDRs L1-3 of SEQ ID NOs: 100-102 of the first binding domain, and CDRs H1-3 of SEQ ID NOs: 190-192 and CDRs L1-3 of SEQ ID NOs: 193-195 of the second binding domain. The CD20- and CD22-targeting antigen binding molecule of any one of claims 1 to 5, comprising a VH region comprising CDR-H1, CDR-H2, and CDR-H3 selected from the group consisting of CDR-L1, CDR-L2, and CDR-L3.

7. The CD20 and CD22 targeted antigen binding molecule of any one of claims 1 to 6, wherein the first binding domain is capable of binding to CD20 and the second binding domain is simultaneously capable of binding to CD22, and wherein CD20 and CD22 are present on the same target cell.

8. The third binding domain comprises: a) CDR H1-3 of SEQ ID NOs: 392-394, and CDR L1-3 of SEQ ID NOs: 395-397; and b) CDRs H1-3 of SEQ ID NOs: 401-403 and CDRs L1-3 of SEQ ID NOs: 404-406 2. The CD20- and CD22-targeting antigen binding molecule of claim 1, comprising a VH region comprising CDR-H1, CDR-H2, and CDR-H3 selected from the group consisting of CDR-L1, CDR-L2, and CDR-L3.

9. The antigen-binding molecule has, in the order from amino to carboxyl, (a) the first domain; (b) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 1-4 and 9-12, or the amino acid sequence of SEQ ID NO: 11; (c) the second domain; (d) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 1-3; and (e) the third domain The CD20 and CD22 targeting antigen binding molecule of any one of claims 1 to 8, comprising:

10. The antigen-binding molecule has, in the order from amino to carboxyl, (f) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 1, 2, 3, 9, 10, 11, and 12; (g) a first polypeptide monomer of the fourth domain; (h) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 5, 6, 7, and 8; and (i) a second polypeptide monomer of the fourth domain. The CD20 and CD22 targeted antigen binding molecule of claim 9, further comprising:

11. 11. The CD20 and CD22 targeted antigen binding molecule of any one of claims 1 to 10, wherein the first binding domain comprises a VH and VL region selected from SEQ ID NO: 64 as VH and SEQ ID NO: 65 as VL, SEQ ID NO: 77 as VH and SEQ ID NO: 78 as VL, SEQ ID NO: 90 as VH and SEQ ID NO: 91 as VL, SEQ ID NO: 103 as VH and SEQ ID NO: 104 as VL, and the second binding domain comprises a VH and VL region selected from SEQ ID NO: 144 as VH and SEQ ID NO: 145 as VL, SEQ ID NO: 157 and 158, SEQ ID NO: 172 and 173, SEQ ID NO: 183 and 184, SEQ ID NO: 196 and 197, SEQ ID NO: 209 and 210, SEQ ID NO: 131 and 132, and SEQ ID NO: 385 and 386, respectively.

12. The CD20 and CD22 targeting antigen binding molecule of any one of claims 1 to 11, wherein the first binding domain comprises an scFv sequence selected from the group consisting of SEQ ID NOs: 66, 79, 92, and 105, and the second binding domain comprises an scFv sequence selected from the group consisting of SEQ ID NOs: 146, 159, 172, 185, 198, 211, 133, 224, and 387.

13. The CD20 and CD22 targeted antigen binding molecule of any one of claims 1 to 12, wherein the antigen binding molecule comprises a first (CD20) and a second (CD22) target binding domain together with a third effector (CD3) binding domain and a fourth domain that confers half-life extension, wherein the three binding domains and the fourth domain linked together have a sequence selected from the group consisting of SEQ ID NOs: 238, 248, 258, 268, 278, 288, 308, 318, 328, 338, 348, 368, and 378.

14. A process for producing a CD20- and CD22-targeted antigen-binding molecule according to any one of claims 1 to 13, comprising culturing a host cell transformed or transfected with a polynucleotide encoding the antigen-binding molecule according to any one of claims 1 to 13 or a vector comprising a polynucleotide encoding the antigen-binding molecule according to any one of claims 1 to 13 under conditions allowing expression of the antigen-binding molecule according to any one of claims 1 to 13, and recovering the produced antigen-binding molecule from the culture.

15. A pharmaceutical composition comprising a CD20- and CD22-targeting antigen binding molecule according to any one of claims 1 to 13.

16. 16. The pharmaceutical composition of claim 15 for use in the treatment of a disease selected from a proliferative disease, a neoplastic disease, a cancer, and an immune disorder.

17. The pharmaceutical composition of claim 16 for use in the treatment of a cancer selected from non-Hodgkin's lymphoma (NHL), non-small cell lung cancer (NSCLC), and colorectal cancer (CRC).

Images