Method for purifying a bispecific antigen-binding polypeptide having an enhanced protein L capture dynamic binding capacity

Abstract

The present invention provides a downstream purification process for the production of bispecific antigen-binding polypeptides, comprising at least the steps of (i) providing a separation resin comprising a polymer matrix portion and a ligand portion, wherein the matrix portion comprises polymethacrylate and has a particle size of approximately 30-60 μm, and the ligand portion comprises recombinant Protein L, wherein the Protein L in the ligand portion is covalently bound to the particles of the matrix portion, (ii) contacting a process fluid containing the bispecific antigen-binding polypeptide with the separation resin, (iii) capturing the bispecific antigen-binding polypeptide by the ligand portion of the separation resin, wherein the bispecific antigen-binding polypeptide reversibly binds to the ligand portion of the separation resin, and the remainder of the process fluid does not bind to the ligand portion of the separation resin, (iv) washing the bound bispecific antigen-binding polypeptide with a wash buffer that does not elute the bispecific antigen-binding polypeptide from the ligand portion, and (v) eluting the bispecific antigen-binding polypeptide from the ligand portion with an elution buffer having a low pH.

Description

[Technical Field]

[0001] The present invention relates to a biotechnology method, and more particularly to a method for the downstream purification of bispecific antigen polypeptides. [Background technology]

[0002] Despite advances in manufacturing, novel protein-based pharmaceuticals require new and optimized manufacturing processes to avoid product quality issues such as protein aggregation. This impacts upstream manufacturing, downstream manufacturing, storage, and application.

[0003] Such novel protein-based pharmaceuticals include, for example, bispecific antigen-binding polypeptides, such as (monoclonal) antibodies. Bispecific polypeptides, such as antibodies, are artificial proteins that can simultaneously bind to two different types of antigens. They are known in several structural forms, and their applications in cancer immunotherapy and drug delivery are currently being explored (Fan, Gaowei; Wang, Zujian; Hao, Mingju; Li, Jinming (2015). “Bispecific antibodies and their applications”. Journal of Hematology & Oncology. 8:130).

[0004] Generally, bispecific antibodies may be IgG-like, i.e., full-length bispecific antibodies, or non-IgG-like bispecific antibodies (which are, for example, not full-length antibodies). Full-length bispecific antibodies typically retain the normal monoclonal antibody (mAb) structure of two Fab arms and one Fc region, except that the two Fab sites bind to different antigens. Non-full-length bispecific antibodies completely lack the Fc region. These include chemically linked Fabs consisting only of Fab regions, as well as various types of bivalent and trivalent single-stranded variable fragments (scFv). There are also fusion proteins that mimic the variable domains of two antibodies. Among these new forms, the most advanced in development appears to be bispecific T-cell engagers (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).

[0005] Bispecific antigen-binding polypeptides, such as BiTE® molecules, are recombinant protein constructs composed of binding domains derived from two flexibly linked antibodies. One binding domain of a BiTE® molecule is specific to a selected tumor-associated surface antigen on target cells, while the second binding domain is specific to CD3, a subunit of the T cell receptor complex on T cells. Due to their special design, BiTE® molecules are uniquely suited to transiently binding T cells to target cells while simultaneously potently activating the intrinsic cytolytic ability of T cells against target cells. Significant further developments of the first-generation BiTE® molecules, deployed in clinics as AMG 103 and AMG 110 (see International Publication No. 99 / 54440 and International Publication No. 2005 / 040220), provided bispecific molecules that bind to a context-independent epitope at the N-terminus of the CD3 ε chain (see International Publication No. 2008 / 119567). The BiTE® molecule that binds to this selected epitope not only does not exhibit interspecies specificity for human and common marmoset (Callithrix jacchus), cotton-top tamarin (Saguinus oedipus), or squirrel monkey (Saimiri sciureus) CD3ε chains, but also recognizes this specific epitope instead of the CD3 conjugate epitopes previously described in bispecific T cell engagement molecules, thus not nonspecifically activating T cells to the same extent as observed with previous generation T cell engagement antibodies. This reduced T cell activation is associated with less or decreased T cell redistribution in patients, which was judged to be a risk of side effects.

[0006] Currently, bispecific antigen-binding polypeptides are typically processed by downstream chromatography purification using affinity resins for antibody fragment purification. When bispecific antibodies are constructs lacking the Fc region necessary for binding to protein A, as is the case with some BiTE® molecules, an alternative ligand is required for their affinity purification. Protein L, isolated from the surface of bacterial species, has been found to bind to immunoglobulins via the light chain of bispecific antigen-binding polypeptides. Such affinity resins typically contain an immunoglobulin-binding recombinant protein L ligand in a rigid, highly fluid agarose matrix, which exhibits strong affinity for the variable region of the antibody's kappa light chain. Such resins are considered suitable for capturing a wide range of antibody fragments, including fragment antibody-binding (Fab), dAb, and single-strand fragment-variable (scFv), and are intended to have high binding capacity, low ligand leakage, and selectivity for a wide range of antibody fragments, thereby preferably reducing process time and resin volume. However, novel complex molecules, such as bispecific antigen-binding polypeptides with scFv form, require specific and tailored downstream purification solutions to fully utilize their potential benefits. Novel bispecific antigen-binding polypeptides with advantageous therapeutic properties will not offer practical benefits if the available purification methods result in, for example, poor monomer content, long purification times, and therefore low overall productivity.

[0007] Therefore, in order to provide sufficient quantities of commercially viable products of a quality that reduces the amount of waste required in downstream processes, there is a need for improved downstream purification methods that improve both product volume and quality, particularly for producing bispecific antibodies. Given the costs of large-scale cell culture processes and the increasing demand for large quantities of low-cost biological products that will be supplied to patients with severe unmet medical needs, new process methods that can gradually improve the production and recovery of recombinant proteins are useful. [Prior art documents]

Patent Documents

[0008]

Patent Document 1

Patent Document 2

Patent Document 3

Non-Patent Documents

[0009]

Non-Patent Document 1

Non-Patent Document 2

Summary of the Invention

Means for Solving the Problems

[0010] Surprisingly, a suitable downstream purification method can be provided that ensures improvement in both the production amount and product quality of bispecific antibodies. Although several materials (including different protein L resins) for the downstream purification of antibody fragments are known, the one most suitable for the production of scFv bispecific antigen-binding polypeptides has not been determined so far.

[0011] Therefore, in one embodiment, a method for purifying a bispecific antigen-binding polypeptide comprising a first domain that binds to a cell surface antigen and a second domain that binds to an extracellular epitope of human and macaque CD3ε chains is envisioned in relation to the present invention, the method comprising the following steps: (a) A step of providing a separation resin comprising a polymer matrix portion and a ligand portion (wherein the matrix portion comprises a polymer, preferably polymethacrylate, and has a particle size of at least 10 μm, preferably at least 20 μm, more preferably about 30 to 60 μm, and the ligand portion comprises recombinant protein L, and the protein L of the ligand portion is covalently bonded to the particles of the matrix portion), (b) A step of contacting a separation resin with a process fluid containing a bispecific antigen-binding polypeptide, (c) A step in which the bispecific antigen-binding polypeptide is captured by the ligand portion of the separation resin (where the bispecific antigen-binding polypeptide reversibly binds to the ligand portion of the separation resin, and the remainder of the process fluid does not bind to the ligand portion of the separation resin), (d) A step of washing the bound bispecific antigen-binding polypeptide with a washing buffer that does not elute the bispecific antigen-binding polypeptide from the ligand moiety, and (e) A step of eluting the bispecific antigen-binding polypeptide from the ligand moiety with an acidic pH elution buffer.

[0012] According to the above-mentioned aspect of the present invention, the matrix portion has a particle size of about 45 μm.

[0013] According to the above-described aspect of the present invention, recombinant protein L comprises a modified B4 domain having an alkali-stable tetrameric ligand having a plurality of coupling sites.

[0014] The method according to claim 1, wherein the recombinant protein L is reversibly bound to the κ light chain outside the antigen-binding site of the bispecific antigen-binding polypeptide.

[0015] According to the above-described aspect of the present invention, the process fluid is passed through the separation resin at least once (purification cycle) to allow contact between the bispecific antigen-binding polypeptide and protein L (residence time) (wherein the residence time of the bispecific antigen-binding polypeptide before elution is at least about 2 minutes, preferably about 2.5 to 4 minutes).

[0016] According to the above-mentioned aspect of the present invention, the washing buffer preferably comprises phosphate-buffered saline (PBS) in a concentration range of 0.01 to 1 times, preferably 3-(N-morpholino)propanesulfonic acid (MOPS) in a concentration range of 0 to 30 mM, preferably NaCl in a concentration range of 50 to 150 mM, preferably Tris in a concentration range of 15 to 35 mM, preferably arginine in a concentration range of 0.25 to 1 M, and preferably acetate in a concentration range of 40 to 60 mM, and the pH is in the range of 5 to 8.

[0017] According to the above-described aspect of the present invention, the elution buffer preferably comprises at least one compound selected from the group consisting of tris in the range of 15 to 35 mM, arginine in the range of 0.25 to 1 M, glycine in the range of 50 to 150 mM, and acetate in the range of 50 to 150 mM, and has a pH in the range of about 3 to 7.5, preferably 3.3 to 4.2.

[0018] According to the above-mentioned aspect of the present invention, the dynamic load capacity is at least 10 mg / ml resin, preferably at least 15 mg / ml resin, and more preferably at least 18 mg / ml resin.

[0019] According to the above-described aspect of the present invention, the elution and binding capacity is at least 7.5 mg / ml resin, preferably at least 9 mg / ml resin, and more preferably 16 mg / ml resin.

[0020] According to the above-described aspect of the present invention, the bispecific antigen-binding polypeptide is a third domain comprising two polypeptide monomers, each comprising a hinge, a CH2 domain, and a CH3 domain, wherein the two polypeptide monomers further comprise a third domain fused to each other via a peptide linker.

[0021] According to the above-described aspect of the present invention, the antigen-binding polypeptide is a single-chain antigen-binding polypeptide.

[0022] According to the above-described aspect of the present invention, the third domain is arranged in the order of amino to carboxyl, Hinge-CH2-CH3-Linker-Hinge-CH2-CH3 Includes.

[0023] According to the above-described aspect of the present invention, each of the polypeptide monomers in 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: 203 to 210.

[0024] According to the above-mentioned aspect of the present invention, each of the polypeptide monomers has an amino acid sequence selected from SEQ ID NOs. 203 to 210.

[0025] According to the above-mentioned aspect of the present invention, the CH2 domain includes intradomain cysteine ​​disulfide crosslinks.

[0026] According to the above-mentioned aspect of the present invention, (i) Does the first domain contain two antibody-variable domains, and the second domain contain two antibody-variable domains? (ii) Does the first domain contain one antibody variable domain and the second domain contain two antibody variable domains? (iii) The first domain contains two antibody variable domains and the second domain contains one antibody variable domain; or (iv) The first domain contains one antibody-variable domain, and the second domain contains one antibody-variable domain.

[0027] According to the above-mentioned aspect of the present invention, the first and second domains are fused to the third domain via a peptide linker.

[0028] According to the above-described aspect of the present invention, the antigen-binding polypeptide is composed of amino and carboxyl molecules in that order. (a) The first domain; (b) Preferably a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 187 to 189; (c) Second domain Includes.

[0029] According to the above-described aspect of the present invention, the antigen-binding polypeptide is composed of amino and carboxyl molecules in that order. (d) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 187, 188, 189, 195, 196, 197, and 198; (e) The 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: 191, 192, 193, and 194; and (g) Second polypeptide monomer of the third domain It also includes.

[0030] According to the above-described aspect of the present invention, the first domain of the antigen-binding polypeptide binds to an epitope of CD33, CD19, BCMA, PSMA, EGFRvIII, MUC17, FLT3, CD70, DLL3, CDH3, or EpCAM, preferably CD33.

[0031] According to the above-described aspect of the present invention, the first binding domain includes a VH region comprising CDR-H1, CDR-H2, and CDR-H3 selected from the following: (a) CDR-H1 as shown in SEQ ID NO: 1, CDR-H2 as shown in SEQ ID NO: 2, CDR-H3 as shown in SEQ ID NO: 3, CDR-L1 as shown in SEQ ID NO: 4, CDR-L2 as shown in SEQ ID NO: 5, and CDR-L3 as shown in SEQ ID NO: 6 (b) CDR-H1 as shown in SEQ ID NO: 29, CDR-H2 as shown in SEQ ID NO: 30, CDR-H3 as shown in SEQ ID NO: 31, CDR-L1 as shown in SEQ ID NO: 34, CDR-L2 as shown in SEQ ID NO: 35, and CDR-L3 as shown in SEQ ID NO: 36 (c) CDR-H1 as shown in SEQ ID NO: 42, CDR-H2 as shown in SEQ ID NO: 43, CDR-H3 as shown in SEQ ID NO: 44, CDR-L1 as shown in SEQ ID NO: 45, CDR-L2 as shown in SEQ ID NO: 46, and CDR-L3 as shown in SEQ ID NO: 47 (d) CDR-H1 as shown in SEQ ID NO: 53, CDR-H2 as shown in SEQ ID NO: 54, CDR-H3 as shown in SEQ ID NO: 55, CDR-L1 as shown in SEQ ID NO: 56, CDR-L2 as shown in SEQ ID NO: 57, and CDR-L3 as shown in SEQ ID NO: 58 (e) CDR-H1 as shown in SEQ ID NO: 65, CDR-H2 as shown in SEQ ID NO: 66, CDR-H3 as shown in SEQ ID NO: 67, CDR-L1 as shown in SEQ ID NO: 68, CDR-L2 as shown in SEQ ID NO: 69, and CDR-L3 as shown in SEQ ID NO: 70 (f) CDR-H1 as shown in SEQ ID NO: 83, CDR-H2 as shown in SEQ ID NO: 84, CDR-H3 as shown in SEQ ID NO: 85, CDR-L1 as shown in SEQ ID NO: 86, CDR-L2 as shown in SEQ ID NO: 87, and CDR-L3 as shown in SEQ ID NO: 88 (g) CDR-H1 as shown in SEQ ID NO: 94, CDR-H2 as shown in SEQ ID NO: 95, CDR-H3 as shown in SEQ ID NO: 96, CDR-L1 as shown in SEQ ID NO: 97, CDR-L2 as shown in SEQ ID NO: 98, and CDR-L3 as shown in SEQ ID NO: 99 (h) CDR-H1 as shown in SEQ ID NO: 105, CDR-H2 as shown in SEQ ID NO: 106, CDR-H3 as shown in SEQ ID NO: 107, CDR-L1 as shown in SEQ ID NO: 109, CDR-L2 as shown in SEQ ID NO: 110, and CDR-L3 as shown in SEQ ID NO: 111 (i) CDR-H1 as shown in SEQ ID NO: 115, CDR-H2 as shown in SEQ ID NO: 116, CDR-H3 as shown in SEQ ID NO: 117, CDR-L1 as shown in SEQ ID NO: 118, CDR-L2 as shown in SEQ ID NO: 119, and CDR-L3 as shown in SEQ ID NO: 120 (j) CDR-H1 as shown in SEQ ID NO: 126, CDR-H2 as shown in SEQ ID NO: 127, CDR-H3 as shown in SEQ ID NO: 128, CDR-L1 as shown in SEQ ID NO: 129, CDR-L2 as shown in SEQ ID NO: 130, and CDR-L3 as shown in SEQ ID NO: 131 (k) CDR-H1 as shown in SEQ ID NO: 137, CDR-H2 as shown in SEQ ID NO: 138, CDR-H3 as shown in SEQ ID NO: 139, CDR-L1 as shown in SEQ ID NO: 140, CDR-L2 as shown in SEQ ID NO: 141, and CDR-L3 as shown in SEQ ID NO: 142 (l) CDR-H1 as shown in SEQ ID NO: 152, CDR-H2 as shown in SEQ ID NO: 153, CDR-H3 as shown in SEQ ID NO: 154, CDR-L1 as shown in SEQ ID NO: 155, CDR-L2 as shown in SEQ ID NO: 156, and CDR-L3 as shown in SEQ ID NO: 157, (m) CDR-H1 as shown in Sequence ID No. 167, CDR-H2 as shown in Sequence ID No. 168, CDR-H3 as shown in Sequence ID No. 169, CDR-L1 as shown in Sequence ID No. 170, CDR-L2 as shown in Sequence ID No. 171, and CDR-L3 as shown in Sequence ID No. 172.

[0032] According to the above-described aspect of the present invention, the antigen-binding polypeptide is composed of amino and carboxyl molecules in that order. (a) The first domain as described above, (b) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs. 187-189, (c) A second domain having 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 as described in International Publication No. 2008 / 119567. Includes.

[0033] According to the above-described aspect of the present invention, the antigen-binding polypeptide is composed of amino and carboxyl molecules in that order. (d) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 187, 188, 189, 195, 196, 197, and 198; (e) A first polypeptide monomer of a third domain having an amino acid sequence selected from the group consisting of SEQ ID NOs. 203-210, (f) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 191, 192, 193, 194 and 195, and (g) A second polypeptide monomer of the third domain having an amino acid sequence selected from the group consisting of SEQ ID NOs. 203-210. It also includes.

[0034] According to the above-described aspect of the present invention, the bispecific antigen-binding polypeptide has an amino acid sequence selected from the group consisting of "bispecific (HLE) molecules" listed in Table 11.

[0035] According to the above-described aspect of the present invention, the pharmaceutical composition comprises a bispecific antigen-binding polypeptide as described herein.

[0036] According to the above-described aspect of the present invention, the bispecific antibody is for use in the prevention, treatment, or improvement of a disease selected from proliferative disorders, neoplastic diseases, cancer, or immunodeficiency.

[0037] A second aspect of the present invention provides a method for improving the yield of a manufacturing process for a bispecific antigen-binding polypeptide, wherein the method according to the first aspect is applied in a downstream process. [Brief explanation of the drawing]

[0038] [Figure 1A] Figure 1 shows four chromatograms with bispecific antigen-binding polypeptide elution characteristics under the following four different Protein L resins in a capture chromatography column: (A) TOYOPEARL® AF-r Protein L-650F resin, (B) GE Kappaselect Protein L resin, (C) GE Lambdaselect Protein L resin, and (D) Kappa XL Protein L resin. In (A), a significant elution peak was obtained after washing, whereas in (B), (C), and (D), no significant elution peak was observed after washing, but load breakthrough occurred prematurely and unfavorably. [Figure 1B] Figure 1 shows four chromatograms with bispecific antigen-binding polypeptide elution characteristics under the following four different Protein L resins in a capture chromatography column: (A) TOYOPEARL® AF-r Protein L-650F resin, (B) GE Kappaselect Protein L resin, (C) GE Lambdaselect Protein L resin, and (D) Kappa XL Protein L resin. In (A), a significant elution peak was obtained after washing, whereas in (B), (C), and (D), no significant elution peak was observed after washing, but load breakthrough occurred prematurely and unfavorably. [Figure 1C]Figure 1 shows four chromatograms with bispecific antigen-binding polypeptide elution characteristics under the following four different Protein L resins in a capture chromatography column: (A) TOYOPEARL® AF-r Protein L-650F resin, (B) GE Kappaselect Protein L resin, (C) GE Lambdaselect Protein L resin, and (D) Kappa XL Protein L resin. In (A), a significant elution peak was obtained after washing, whereas in (B), (C), and (D), no significant elution peak was observed after washing, but load breakthrough occurred prematurely and unfavorably. [Figure 1D] Figure 1 shows four chromatograms with bispecific antigen-binding polypeptide elution characteristics under the following four different Protein L resins in a capture chromatography column: (A) TOYOPEARL® AF-r Protein L-650F resin, (B) GE Kappaselect Protein L resin, (C) GE Lambdaselect Protein L resin, and (D) Kappa XL Protein L resin. In (A), a significant elution peak was obtained after washing, whereas in (B), (C), and (D), no significant elution peak was observed after washing, but load breakthrough occurred prematurely and unfavorably. [Figure 2] Figure 2 shows a comparison of the binding capabilities of conventional Capto L resin [gray bar] and TOYOPEARL® AF-r protein L-650F [black bar] to CD33×CD3 bispecific antigen-binding polypeptides during the loading and elution phases. [Modes for carrying out the invention]

[0039] A downstream purification method for producing therapeutic proteins, particularly scFv bispecific antigen-binding polypeptides, is provided herein. The present invention envisions adapting the downstream process to the specific requirements of producing bispecific antibodies. The downstream purification method not only contributes to increased productivity and reduced space requirements compared to standard purification using Protein L-packed columns known in the art, but also to further enhance productivity as a chromatographic capture step in the downstream process, and is expected to result in bispecific antibodies with higher product quality, i.e., less aggregation of bispecific antibodies in terms of higher monomer content, with respect to the use of Protein L-packed columns such as Capto® L.

[0040] It has been found that using a specific chromatographic capture step in the downstream purification according to the present invention, i.e., using a recombinant protein L ligand covalently bound to a base matrix preferably consisting of polymethacrylate having a particle size of about 30-60 μm, preferably results in a significant increase in dynamic binding capacity, leading to higher load and a decrease in the collected cell culture pool volume. This, overall, reduces equipment fit.

[0041] For example, high elution-binding capacities exceeding those of typical 10 g / L-filled resins are achieved, such as 12.7 g / L-filled resins using TOYOPEARL® AF-r Protein L-650F resin, resulting in at least a 2x, preferably at least 3x, or even 4x improvement in elution-binding capacity compared to standard affinity resins Capto® L, while the overall yield remained within the same range as with current processes. What is remarkable from a cutting-edge technological standpoint is that several other advantages are achievable with such improvements in (elution) binding capacity. For example, the use of a chromatographic capture step in a downstream process using a resin with a 4x improvement in (elution) binding capacity compared to Capto L, such as TOYOPEARL AF-r Protein L-650F, surprisingly results in a 6x reduction in the number of purification cycles required for a given volume of process fluid, i.e., a reduction in a given volume of collected cell culture medium from, for example, about 12 purification cycles to 2 purification cycles for CD33×CD3 bispecific antigen-binding polypeptides as described herein. As those skilled in the art will understand, such a significant reduction in the required purification cycle reduces the amount of time, space, and energy required to process a given amount of process fluid containing the bispecific antigen-binding polypeptide to be purified.

[0042] In relation to the present invention, the increase in efficiency corresponds to a significant reduction in the residence time of the bispecific antigen-binding polypeptide on the resin in one purification cycle. Furthermore, the increase in effectiveness is related to a reduction in the time required to load with maximum binding capacity at a given residence time. For example, with a residence time of 3 minutes compared to 5 minutes with conventional Capto L, the time required to load with maximum binding capacity is reduced from approximately 7 hours to approximately 4 hours.

[0043] In relation to the present invention, one purification cycle corresponds to the time span during which the target protein is loaded onto the resin in the separation column and resides on the resin, taking into account the time required for washing and the time required for protein elution. Typically, loading takes several hours, preferably 7 hours or less, more preferably 5 hours or less, while the residence time is preferably as short as 2 minutes, or can last for 3, 4, or 5 minutes. Longer protein residence times, and therefore purification cycles, are rare and undesirable in relation to the present invention. For example, the loading time required to achieve a higher loading factor targeting 18 g / L for a BCMAxCD3 bispecific antigen-binding polypeptide as described herein is typically up to 7 hours. Typically, loading is the greatest time factor of the cycle. Therefore, the cycle time depends on the time required to load at the maximum binding capacity, which is typically 80-90% of the dynamic binding capacity of each resin.

[0044] In relation to the present invention, residence time is calculated by dividing the column bed height by the linear flow rate. For example, if the residence time is 3 minutes, the loading time will be fast because the protein stays in the column for a short time. Alternatively, if the residence time is 6 minutes, the linear flow rate [cm / h] is halved at the same bed height, so the loading time will be longer. Therefore, longer residence times are undesirable in relation to the present invention. However, if the target loading rate is very high and the maximum binding capacity is also high, longer loading times are intended in relation to the present invention. The present invention aims to load more bispecific antigen-binding polypeptides with a short processing time or short residence time.

[0045] In relation to the present invention, one purification cycle typically includes equilibration, loading, at least one washing step including the same washing 1 as equilibration and optionally washing 2, elution, exfoliation, washing, optionally regeneration between cycles, but usually only after the last cycle of the batch, and storage.

[0046] In relation to the present invention, it is understood that proteins A and G bind to the Fc region of the heavy chain, while protein L binds to the κ-light chain outside the antigen-binding site. Structural studies reveal that distinct motifs are responsible for binding: domains E, D, A, B, and C in protein A; domains C1, D1, and C2 in protein G; and domains B1, B2, B3, and B4 in protein L.

[0047] In relation to the present invention, protein L in which the B4 domain has been modified, such as TOYOPEARL AF-r protein L-650F, is preferred. Typically, TOYOPEARL AF-r protein L-650F contains a matrix comprising a polymer, preferably polymethacrylate, and preferably having a particle size of about 30-60 μm, to which ligand protein L, with a modified B4 domain, is covalently bound.

[0048] In relation to the present invention, the target load (g / L filling resin) is understood to be at least 80%, preferably at least 90%, of the dynamic binding capacity, which, when actually performed on the resin, typically does not result in an initial load breakthrough per cycle. This is preferable in relation to the present invention. An initial load breakthrough is understood herein as a phenomenon observed when the resin can no longer hold a predetermined set load rate for molecules to be bound to it, such as antigen-binding polypeptides, and the molecules are present in the liquid passing through the resin, such as a flow-through load sample, without binding to the resin. Typically, a load breakthrough occurs when the concentration of load molecules, such as bispecific antigen-binding polypeptides, in the flow-through pool becomes equal to the concentration of the feed solution.

[0049] In relation to the present invention, elution binding capacity (g / L of the packed resin) is understood as the maximum amount of the target molecule (e.g., bispecific antigen-binding polypeptide) typically recovered in the eluted elution pool as a result of using a buffer whose affinity for the ligand is typically higher than that of the target molecule. Elution binding capacity is also typically expressed as a recovery yield percentage of the resin, generally an affinity resin, and is calculated as the percentage of total antibody (mass) filtered and recovered per unit volume of the packed resin. Theoretically, elution binding capacity should be equal to the load binding capacity, but generally it is less than the load binding capacity because it depends on the strength of the elution buffer used and not all proteins loaded into the resin are eluted.

[0050] In the context of this invention, column ID (cm) is understood as the inner diameter of the column. A larger diameter allows more process fluid to pass through within a given time frame.

[0051] In relation to the present invention, "cell culture" or "culture" means the growth and proliferation of cells outside of a multicellular organism or tissue. Suitable culture conditions for mammalian cells are known in the art. See, for example, *Animal cell culture: A Practical Approach*, D. Rickwood, ed., Oxford University Press, New York (1992). Mammalian cells can be cultured in suspension or attached to a solid substrate.

[0052] The term “mammalian cell” means any cell from or derived from any mammal (e.g., human, hamster, mouse, green monkey, rat, pig, cow, or rabbit). For example, a mammalian cell may be an immortalized cell. In some embodiments, the mammalian cell is a differentiated cell. In some embodiments, the mammalian cell is an undifferentiated cell. Non-limiting examples of mammalian cells are described herein. A preferred type of mammalian cell in connection with the present invention is a GS-KO cell. Further examples of mammalian cells are known in the art.

[0053] As used herein, the term “cell culturing medium” (also known as “culture medium,” “cell culture media,” or “tissue culture medium”) refers to any nutrient solution used for growing cells, e.g., animal or mammalian cells, and generally provides at least one of the following: an energy source (usually in the form of carbohydrates such as glucose); all essential amino acids, and generally one or more of the 20 basic amino acids and cysteine; vitamins and / or other organic compounds typically required in low concentrations; lipids or free fatty acids; and trace elements, e.g., inorganic compounds or naturally occurring elements, typically required in very low concentrations, usually in the micromolar concentration range.

[0054] Cell culture media include, but are not limited to, those commonly used in and / or known for use in any cell culture process, such as cell batches, expansion batches, fed batches, and / or perfusion or continuous culture.

[0055] "Perfused" cell culture medium or feed medium refers to a cell culture medium commonly used in cell culture that is maintained by perfusion or continuous culture and is sufficiently complete to aid cell culture during this process. Perfused cell culture medium formulations may be richer or more concentrated than basic cell culture medium formulations to accommodate the method used to remove used medium. Perfused cell culture mediums may be used in both the growth and production phases.

[0056] The term "0.5 × quantity" means approximately 50% of the quantity. The term "0.6 × quantity" means approximately 60% of the quantity. Similarly, 0.7 ×, 0.8 ×, 0.9 ×, and 1.0 × mean approximately 70%, 80%, 90%, and 100% of the quantity, respectively.

[0057] The term "culture" or "cell culture" refers to the maintenance or proliferation of mammalian cells under a controlled set of physical conditions.

[0058] The term "mammalian cell culture" refers to a liquid culture medium containing multiple mammalian cells maintained or grown under a controlled set of physical conditions.

[0059] The term "liquid culture medium" refers to a liquid containing sufficient nutrients to grow or proliferate cells (e.g., mammalian cells) in vitro. For example, a liquid culture medium may contain one or more of the following: amino acids (e.g., 20 amino acids), purines (e.g., hypoxanthine), pyrimidines (e.g., thymidine), choline, inositol, thiamine, folic acid, biotin, calcium, niacinamide, pyridoxine, riboflavin, thymidine, cyanocobalamin, pyruvate, lipoic acid, magnesium, glucose, sodium, potassium, iron, copper, zinc, and sodium bicarbonate. In some embodiments, the liquid culture medium may contain mammalian serum. In some embodiments, the liquid culture medium does not contain mammalian serum or other extracts (standard liquid culture medium). In some embodiments, the liquid culture medium may contain trace metals, mammalian growth hormones, and / or mammalian growth factors. Another example of a liquid culture medium is a minimal medium (e.g., a medium containing only inorganic salts, a carbon source, and water). Non-limiting examples of liquid culture media are described herein. Further examples of liquid culture media are known in the art and are commercially available. Liquid culture media may contain mammalian cells of any density. For example, in the use herein, the volume of liquid culture medium removed from the bioreactor may be substantially free of mammalian cells.

[0060] The term “continuous process” means a process through which a liquid is continuously supplied via at least a portion of the system. For example, in any of the exemplary continuous biological manufacturing systems described herein, a liquid culture medium containing recombinant therapeutic protein is continuously supplied to the system while the system is in operation, and therapeutic proteinogens are supplied from the system.

[0061] The term "clipping" usually refers to the partial cleavage of an expressed protein through proteolysis.

[0062] The term "decomposition" generally refers to the breakdown of a larger entity, such as a peptide or protein, into at least two smaller entities, one of which may be significantly larger than the other.

[0063] The term "deamide" typically refers to any chemical reaction in which the amide functional group of the side chain of an amino acid, such as asparagine or glutamine, is removed or converted to another functional group. Typically, asparagine is converted to aspartic acid or isoaspartic acid.

[0064] The term "aggregation" generally refers to direct intermolecular attractive forces, such as van der Waals forces or chemical bonds. In particular, aggregation is understood as the accumulation and clumping of proteins. Aggregates may include amorphous aggregates, oligomers, and amyloid fibrils, and are typically referred to as high molecular weight (HMW) species, i.e., molecules having a higher molecular weight than the pure product molecules, which are typically non-aggregated molecules also called low molecular weight (LMW) species or monomers as referred herein.

[0065] Acidic species are understood herein to typically include variants commonly observed when analyzing antibodies by charge-based separation techniques such as isoelectric focusing (IEF) gel electrophoresis, capillary isoelectric focusing (cIEF) gel electrophoresis, cation exchange chromatography (CEX), and anion exchange chromatography (AEX). These variants are referred to as acidic species or basic species in comparison to the major species. When antibodies are analyzed using IEF-based methods, acidic species are typically variants with a low apparent pI, while basic species are variants with a high apparent pI.

[0066] The term "residence time" typically refers to the time a particular product molecule is present in a bioreactor, that is, the time from its creation through biotechnology to its separation from the bioreactor's lumen.

[0067] "Product quality" is typically evaluated by the presence or absence of clipping, degradation, deamide and / or aggregation. For example, products (molecules) with a percentile content of HMW species of less than 40%, preferably less than 35%, or even less than 30%, less than 25%, or less than 20% may be considered to have preferred product quality. Preferred product quality is also associated with a substantial absence of persistent host cell proteins (HCPs), as well as a substantial absence of clipping, degradation, and deamide, or a significant reduction in HCP concentration, clipping, degradation, and / or deamide compared to products produced by processes other than those of the present invention, such as fed-batch processes. Methods known in the art for evaluating product quality in relation to the present invention include cation exchange high-performance chromatography (CEX-HPLC) for charge variant analysis, trypsin peptide mapping for chemical modifications, host cell protein (HCP) ELISA, reductive capillary electrophoresis-sodium dodecyl sulfate (RCE-SDS), and size exclusion high-performance liquid chromatography (SE-HPLC).

[0068] The term “product” refers to “secreted protein” or “secreted recombinant protein,” meaning a protein (e.g., recombinant protein) that originally contains at least one secretory signal sequence when translated within a mammalian cell and is secreted at least partially into the extracellular space (e.g., liquid culture medium) via enzymatic cleavage of the secretory signal sequence within the mammalian cell. Those skilled in the art will understand that a “secreted” protein does not need to be completely dissociated from the cell to be considered a secreted protein.

[0069] The term "polypeptide" is understood herein to mean an organic polymer comprising at least one continuous, unbranched amino acid chain. In connection with the present invention, polypeptides comprising two or more amino acids are also assumed. The amino acid chain of a polypeptide typically contains at least 50 amino acids, preferably at least 100, 200, 300, 400, or 500 amino acids. In connection with the present invention, it is also assumed that the amino acid chain of a polymer is linked to entities that are not composed of amino acids.

[0070] The term “antigen-binding polypeptide” according to the present invention is preferably a polypeptide that binds immunospecifically to its target or antigen. It typically includes or comprises domains that include the heavy chain variable region (VH) and / or light chain variable region (VL) of an antibody. The polypeptide according to the present invention includes the minimum structural requirements of an antibody that enable immunospecific target binding. These minimum requirements can be defined, for example, by the presence of at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VH region), preferably all six CDRs. Thus, a T cell engagement polypeptide may be characterized by the presence of three or six CDRs in one or both binding domains, and those skilled in the art will know where (in what order) those CDRs are located within the binding domain.

[0071] The term "bispecific antigen-binding polypeptide product" encompasses bispecific antibodies, such as full-length antibodies and fragments thereof, including IgG-based antibodies, and is generally referred to herein as bispecific antigen-binding polypeptide.

[0072] Alternatively, in relation to the present invention, an antigen-binding polypeptide such as an “antibody construct” refers to a molecule whose structure and / or function is based on the structure and / or function of an antibody, for example, a full-length or full-length immunoglobulin molecule (usually composed of two untruncate heavy chains and two light chains), and / or extracted from the variable heavy chain (VH) and / or variable light chain (VL) domains of an antibody or a fragment thereof. Thus, an antigen-binding polypeptide can bind to its specific target or antigen. Furthermore, in this specification, the domain that binds to the binding partner according to the present invention is understood to be the binding domain of the antigen-binding polypeptide according to the present invention. Typically, the binding domain according to the present invention includes the minimum structural requirements of the antibody that enable target binding. These minimum requirements can be defined, for example, by the presence of at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VH region), preferably all six CDRs. Alternative methods for defining the minimum structural requirements of an antibody include defining the antibody's epitope within the structure of a specific target, or the protein domain of the target protein containing the epitope region (epitope cluster), or by referencing a specific antibody that competes with the epitope of the defined antibody. Antibodies on which the construct according to the present invention is based include, for example, monoclonal antibodies, recombinant antibodies, chimeric antibodies, deimmunized antibodies, humanized antibodies, and human antibodies.

[0073] The binding domain of the antigen-binding polypeptide according to the present invention may include, for example, the CDRs of the group mentioned above. Preferably, these CDRs are contained within the framework of the antibody light chain variable region (VL) and the antibody heavy chain variable region (VH), but do not necessarily contain both. The Fd fragment, for example, has two VH regions and often retains the antigen-binding function of a portion of the intact antigen-binding domain. Further examples of antibody fragments, antibody variants, or binding domains include: (1) Fab fragments, which are monovalent fragments having VL, VH, CL, and CH1 domains; (2) F(ab')2 fragments, which are bivalent fragments having two Fab fragments linked by disulfide crosslinks in a hinge region; (3) Fd fragments having two VH and CH1 domains; (4) Fv fragments having VL and VH domains in one arm of the antibody; (5) dAb fragments having a VH domain (Ward et al., (1989) Nature 341:544-546); (6) isolated complementarity-determining regions (CDRs); and (7) single-stranded Fv (scFv), the latter of which is preferred (e.g., those derived from scFV libraries). Examples of embodiments of the antigen-binding polypeptide according to the present invention are described, for example, in International Publication No. 00 / 006605, International Publication No. 2005 / 040220, International Publication No. 2008 / 119567, International Publication No. 2010 / 037838, International Publication No. 2013 / 026837, International Publication No. 2013 / 026833, U.S. Patent Application Publication No. 2014 / 0308285, U.S. Patent Application Publication No. 2014 / 0302037, International Publication No. 2014 / 144722, International Publication No. 2014 / 151910, and International Publication No. 2015 / 048272.

[0074] 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 “rIgG” (“half-antibody”). The antigen-binding polypeptide according to 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, diabody, single-chain diabody, tandem diabody (Tandab's), tandem di-scFv, tandem tri-scFv, “multibody,” such as triabody or tetrabody, and single-domain antibodies, such as nanobody, or single variable-domain antibodies containing 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.

[0075] As used herein, the terms “single-chain Fv,” “single-chain antibody,” or “scFv” refer to a single polypeptide chain antibody fragment that contains variable regions derived from both the heavy and light chains but lacks a constant region. Typically, single-chain antibodies further include a polypeptide linker between the VH and VL domains, enabling the formation of a desired structure that allows binding to an antigen. 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 for producing single-chain antibodies are known, including, for example, the methods described in U.S. Patent Nos. 4,694,778 and 5,260,203; International Publication No. 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; and Skerra et al. (1988) Science 242:1038-1041. In certain embodiments, single-chain antibodies may also be bispecific, multispecific, human and / or humanized, and / or synthetic.

[0076] Furthermore, the definition of the term "antigen-binding polypeptide" includes monovalent, divalent, and polyvalent constructs, and therefore bispecific constructs that specifically bind to only two antigenic structures, as well as polyspecific constructs that specifically bind to three or more antigenic structures, e.g., three, four, or more, through different binding domains. In addition, the definition of the term "antigen-binding polypeptide" includes molecules consisting of only one polypeptide chain, and molecules consisting of two or more polypeptide chains that may be identical (homodimer, homotrimer, or homooligomer) or different (heterodimer, heterotrimer, or heterooligomer). Examples of the antibodies and variants or derivatives identified above are described, in particular, 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.

[0077] As used herein, the term “bispecificity” means an antigen-binding polypeptide that is “at least bispecific,” i.e., it comprises at least a first binding domain and a second binding domain, where the first binding domain binds to one antigen or target (e.g., a surface antigen of a target cell) and the second binding domain binds to another antigen or target (e.g., CD3). Thus, the antigen-binding polypeptide according to the present invention possesses specificity for at least two different antigens or targets. For example, the first domain preferably does not bind to one or more extracellular epitopes of CD3ε of the species described herein. The term “surface antigen of a target cell” means an antigenic structure expressed by a cell and present on its cell surface so that the antigen-binding polypeptide described herein can access it. 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 polypeptide" in this invention also includes multispecific antigen-binding polypeptides, such as triplicate antigen-binding polypeptides containing three binding domains, or constructs having four or more (e.g., four, five, etc.) specificities.

[0078] The T cell engagement antigen-binding polypeptide according to the present invention is preferably bispecific, which is understood herein to typically include one domain that binds to at least one target antigen and another domain that binds to CD3. Therefore, it does not exist in nature, and its function differs significantly from that of natural products. Accordingly, the polypeptide according to the present invention is an artificial "hybrid" polypeptide comprising at least two distinct binding domains having different specificities, and is therefore bispecific. Bispecific antigen-binding polypeptides can be produced by various methods, including hybridoma fusion or Fab' fragment linking. See, for example, Songsivilai & Lachmann, Clin. Exp. Immunol. 79:315-321 (1990).

[0079] At least two binding domains and variable domains (VH / VL) of the antigen-binding polypeptide of the present invention may or may not include a peptide linker (spacer peptide). According to the present invention, the term "peptide linker" includes an amino acid sequence that links the amino acid sequences of one (variable and / or binding) domain and another (variable and / or binding) domain of the antigen-binding polypeptide of the present invention. A peptide linker may also be used to fuse a third domain to other domains of the antigen-binding polypeptide of the present invention. An essential technical feature of such a peptide linker is that it does not contain any polymerization activity. Suitable peptide linkers include those described in U.S. Patent Nos. 4,751,180 and 4,935,233 or International Publication No. 88 / 09344. Peptide linkers may also be used to bind other domains, modules, or regions (such as half-life extension domains) to the antigen-binding polypeptide of the present invention.

[0080] The antigen-binding polypeptide of the present invention is preferably an "in vitro-produced antigen-binding polypeptide." This term refers to an antigen-binding polypeptide as defined above, in which all or part of the variable region (e.g., at least one CDR) is produced by selecting non-immune cells, e.g., in vitro phage display, protein chip, or any other method that allows the candidate sequence to be tested for antigen-binding ability. Therefore, this term preferably excludes sequences produced solely by genomic rearrangement in animal immune cells. A "recombinant antibody" is an antibody produced by the use of recombinant DNA technology or genetic engineering.

[0081] The term “monoclonal antibody” (mAb), or monoclonal antibodies derived from antigen-binding polypeptides as used herein, refers to antibodies obtained from a substantially homogeneous population of antibodies. That is, the individual antibodies constituting the population are identical except for possible naturally occurring mutations and / or post-translational modifications (e.g., isomerization, amidation) that may be present in small amounts. Monoclonal antibodies are highly specific and are induced to a single antigenic site or determinant on an antigen, in contrast to conventional (polyclonal) antibody preparations, which typically contain different antibodies induced to different determinants (or epitopes). In addition to their specificity, monoclonal antibodies have the advantage of being synthesized by hybridoma culture, thus avoiding contamination with other immunoglobulins. The modifier “monoclonal” indicates the characteristic of an antibody to be obtained from a substantially homogeneous population of antibodies and should not be interpreted as requiring antibody production by any particular method.

[0082] Any technique that yields antibodies produced by continuous cell line culture can be used to prepare monoclonal antibodies. For example, the monoclonal antibodies used may be produced by the hybridoma method first described by Koehler et al., Nature, 256:495 (1975), or by the recombinant DNA method (see, for example, U.S. Patent No. 4,816,567). Further examples of techniques for producing human monoclonal antibodies include trioma techniques, human B-cell hybridoma techniques (Kozbor, Immunology Today 4 (1983), 72), and EBV-hybridoma techniques (Cole et al., Monoclonal Antibodies and Cancer Therapy, Alan R. Liss, Inc. (1985), 77-96).

[0083] Next, hybridomas can be screened using standard methods such as enzyme-linked immunosorbent assay (ELISA) and surface plasmon resonance (BIACORE®) analysis to identify one or more hybridomas that produce antibodies that specifically bind to a particular antigen. Any form of the relevant antigen can be used as an immunogen, for example, recombinant antigen, its naturally occurring form, any variant or fragment thereof, and its antigenic peptide. Surface plasmon resonance, as employed in the BIAcore system, can be used to enhance the efficiency of phage antibody binding to epitopes of surface antigens on target cells (Schier, Human Antibodies Hybridomas 7(1996), 97-105; Malmborg, J.Immunol.Methods 183(1995), 7-13).

[0084] Another exemplary method for producing monoclonal antibodies involves screening protein expression libraries, such as phage display or ribosome display libraries. Phage display is described, for example, in Ladner et al., U.S. Patent 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).

[0085] In addition to using display libraries, non-human animals, such as rodents (mice, hamsters, rabbits, or rats), can be immunized using relevant antigens. In one embodiment, the non-human animal contains at least a portion of the human immunoglobulin gene. For example, a mouse strain lacking mouse antibody production can be modified using a large fragment of the human Ig (immunoglobulin) locus. Using hybridoma technology, antigen-specific monoclonal antibodies derived from genes with desired specificity can be produced and selected. See, for example, XENOMOUSE®, Green et al. (1994) Nature Genetics 7:13-21, U.S. Patent Application Publication No. 2003 / 0070185, International Publication No. 96 / 34096 and International Publication No. 96 / 33735.

[0086] Monoclonal antibodies can also be modified after being obtained from non-human animals using recombinant DNA technologies known in the art, such as humanization, deimmunization, and chimerization. Examples of modified antigen-binding polypeptides include humanized variants of non-human antibodies, "affinity-mature" antibodies (see, e.g., Hawkins et al. J.Mol.Biol. 254, 889-896 (1992) and Lowman et al., Biochemistry 30, 10832-10837 (1991)) and antibody variants with altered effector function (see, e.g., U.S. Patent No. 5,648,260, Kontermann and Duebel (2010) and Little (2009) cited above).

[0087] In immunology, affinity maturation is the process by which B cells produce antibodies with increased affinity for an antigen during an immune response. Repeated exposure to the same antigen leads the host to produce antibodies with progressively increasing affinity. Similar to natural prototypes, in vitro affinity maturation is based on the principles of mutation and selection. Antibodies, antigen-binding polypeptides, and antibody fragments are optimized using in vitro affinity maturation. Random mutations within the CDR are introduced using radiation, chemical mutagens, or error-prone PCR. In addition, genetic diversity can be increased by chain shuffling. Two or three rounds of mutation and selection using display methods such as phage display typically yield antibody fragments with affinity in the low nanomolar range.

[0088] A preferred type of amino acid substitution variant of antigen-binding polypeptide involves substitution of one or more residues in the hypervariable region of the parent antibody (e.g., a humanized antibody or a human antibody). Generally, variants selected for further development will have improved biological properties compared 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 produce all possible amino acid substitutions at each site. The antibody variants thus generated are presented in a monovalent form from filamentous phage particles as a fusion with the M13 gene III product 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 significantly contribute to antigen binding. Alternatively, or in addition, it may be beneficial to analyze the crystal structure of the antigen-antibody complex to identify the contact sites between the binding domain and, for example, the surface antigen of a human target cell. Such contact residues and adjacent residues are candidates for substitution by the techniques detailed herein. After generating such variants, a panel of variants may be screened as described herein, and antibodies exhibiting superior properties in one or more relevant assays may be selected for further development.

[0089] The monoclonal antibodies and antigen-binding polypeptides of the present invention include, in particular, “chimeric” antibodies (immunoglobulins) in which a portion of the heavy chain and / or light chain is derived from a particular species or is identical or homologous to a corresponding sequence in an antibody belonging to a particular class or subclass of antibodies, while the remainder of the chain is identical or homologous to a corresponding sequence in an antibody belonging to another class or subclass of antibodies, and in fragments of such antibodies, insofar as it exhibits the desired biological activity (U.S. Patent No. 4,816,567; Morrison et al., Proc. Natl. Acad. Sci. USA, 81:6851-6855 (1984)). The chimeric antibodies of interest as used herein include “primatized” antibodies containing a variable domain antigen-binding sequence and a human constant region sequence derived from a non-human primate (e.g., Old World monkeys, apes, etc.). Various methods for producing chimeric antibodies are described. For example, see Morrison et al., Proc. Natl. Acad. Sci USA 81:6851, 1985; Takeda et al., Nature 314:452, 1985; Cabilly et al., U.S. Patent No. 4,816,567; Boss et al., U.S. Patent No. 4,816,397; Tanaguchi et al., European Patent No. 0171496; European Patent No. 0173494; and British Patent No. 2177096.

[0090] Antibodies, antigen-binding polypeptides, antibody fragments, or antibody variants can also be modified by specific deletion of human T cell epitopes (a method called "deimmunization") by the methods disclosed in the examples of International Publication No. 98 / 52976 or International Publication No. 00 / 34317. Briefly, for peptides that bind to MHC class II, the heavy and light chain variable domains of antibodies can be analyzed. These peptides correspond to potential T cell epitopes (as defined in International Publication No. 98 / 52976 and International Publication No. 00 / 34317). For the detection of potential T cell epitopes, a computer modeling technique called "peptide threading" can be applied, as described in International Publication No. 98 / 52976 and International Publication No. 00 / 34317, and in addition, motifs present in VH and VL sequences can be searched in a database of human MHC class II-binding peptides. These motifs bind to any of the 18 major MHC class II DR allotypes and thus become potential T cell epitopes. Detected potential T cell epitopes can be removed by substituting a small number of amino acid residues within the variable domain, or preferably by a single amino acid substitution. Conservative substitutions are typically made. In many, though not all, 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 catalog provides a comprehensive overview of human immunoglobulin variable region sequences (edited by Tomlinson, LA. et al., MRC Centre for Protein Engineering, Cambridge, UK). These sequences can be used as a source of human sequences, for example, for framework regions and CDRs.For example, the consensus human framework area described in U.S. Patent No. 6,300,064 can also be used.

[0091] A “humanized” antibody, antigen-binding polypeptide, variant, or fragment thereof (Fv, Fab, Fab', F(ab')2, or other antigen-binding subsequences of an antibody) is an antibody or immunoglobulin that is predominantly human, containing minimal sequences derived from non-human immunoglobulin. In most cases, a humanized antibody is a human immunoglobulin (recipient antibody) in which residues from the recipient’s hypervariable region (also known as the CDR) are replaced with residues from the hypervariable region of a non-human species (e.g., rodents) such as mouse, rat, hamster, or rabbit (donor antibody) that have the desired specificity, affinity, and capability. In some cases, Fv framework region (FR) residues of the human immunoglobulin are replaced with corresponding non-human residues. Furthermore, as used herein, “humanized antibody” may also contain residues not found in either the recipient antibody or the donor antibody. These modifications are made to further refine and optimize the performance of the antibody. A humanized antibody may also contain the immunoglobulin constant region (Fc), usually at least a portion of the constant region of 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).

[0092] Humanized antibodies or fragments thereof can be produced by substituting the sequence of the Fv variable domain, which is not directly involved in antigen binding, with an equivalent sequence derived from the human Fv variable domain. Exemplary methods for producing humanized antibodies or fragments thereof are provided in Morrison (1985) Science 229:1202-1207; Oi et al. (1986) BioTechniques 4:214; and U.S. Patents No. 5,585,089; U.S. Patents No. 5,693,761; U.S. Patents No. 5,693,762; U.S. Patents No. 5,859,205; and U.S. Patents No. 6,407,213. These methods involve isolating, manipulating, and expressing a nucleic acid sequence encoding all or part of the immunoglobulin Fv variable domain derived from at least one of the heavy or light chains. Such nucleic acids can be obtained from hybridomas and other sources that produce antibodies against a given target as described above. Next, recombinant DNA encoding a humanized antibody molecule can be cloned into a suitable expression vector.

[0093] Humanized antibodies can be prepared using transgenic animals, such as mice, that express human heavy and light chain genes but cannot express endogenous mouse immunoglobulin heavy and light chain genes. Winter describes an exemplary CDR grafting method that can be used for the preparation of the humanized antibodies described herein (U.S. Patent No. 5,225,539). All of the CDRs of a particular human antibody may be replaced with at least a portion of non-human CDRs, or only a portion of the CDRs may be replaced with non-human CDRs. Only the number of CDRs required for the humanized antibody to bind to a given antigen needs to be replaced.

[0094] Humanized antibodies can be optimized by introducing conservative substitutions, consensus sequence substitutions, germline substitutions, and / or reverse mutations. Such modified immunoglobulin molecules can be prepared 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 European Patent No. 239400).

[0095] The terms “human antibody,” “human antigen-binding polypeptide,” and “human binding domain” include antibodies, antigen-binding polypeptides, and binding domains having antibody regions such as variable regions and constant regions or domains that substantially correspond to human germline immunoglobulin sequences known in the art, including those described by Kabat et al. (1991) (cited above). The human antibody, antigen-binding polypeptide, or binding domain of the present invention may contain amino acid residues not encoded by human germline immunoglobulin sequences (e.g., mutations introduced by in vitro random mutagenesis or site-directed mutagenesis or in vivo somatic mutation), for example, in the CDR, particularly CDR3. The human antibody, antigen-binding polypeptide, or binding domain may have at least one, two, three, four, five, or more positions substituted with amino acid residues not encoded by human germline immunoglobulin sequences. However, as used herein, the definitions of human antibody, antigen-binding polypeptide, and binding domain also intend to include “fully human antibody” which contains only the human sequence of an antibody that is not artificially and / or genetically modified, and which can be obtained using technologies or systems such as Xenomouse. Preferably, “fully human antibody” does not contain amino acid residues not encoded by a human germline immunoglobulin sequence.

[0096] In some embodiments, the antigen-binding polypeptide of the present invention is an “isolated” or “substantially pure” antigen-binding polypeptide. “Isolated” or “substantially pure,” as used in the description of antigen-binding polypeptides disclosed herein, means an antigen-binding polypeptide identified, separated, and / or recovered from components of its production environment. Preferably, the antigen-binding polypeptide is unrelated or substantially unrelated to all other components from its production environment. Contaminants in its production environment, such as components arising from recombinant transfected cells, are typically materials that interfere with the diagnostic or therapeutic use of the polypeptide, and may include enzymes, hormones, and other proteinaceous or non-proteinaceous solutes. An antigen-binding polypeptide may constitute, for example, at least about 5% by weight or at least about 50% by weight of the total protein in a given sample. An isolated protein is understood to constitute 5% to 99.9% by weight of the total protein content, depending on the environment. Polypeptides can be produced at significantly higher concentrations by the use of inducible promoters or high-expression promoters, so that polypeptides are produced at high concentration levels. This definition includes the production of antigen-binding polypeptides in a variety of organisms and / or host cells known in the art. In preferred embodiments, the antigen-binding polypeptide is purified (1) using a spinning cup sequencer to a degree sufficient to obtain at least 15 residues of the N-terminal or internal amino acid sequence, or (2) to a degree of homogeneity by SDS-PAGE under non-reducing or reducing conditions using Coomassie blue or preferably silver staining. However, the isolated antigen-binding polypeptide is usually prepared by at least one purification step.

[0097] In relation to the present invention, the term "binding domain" is considered to be a domain that (specifically) binds to / interacts with / recognizes a given target epitope or target site on a target molecule (antigen), for example, CD33 and CD3, respectively. The structure and function of the first binding domain (e.g., CD33 recognition), preferably the structure and / or function of the second binding domain (e.g., CD3 recognition), are based on the structure and / or function of the antibody, for example, the full-length or full immunoglobulin molecule, and / or extracted from the variable heavy chain (VH) and / or variable light chain (VL) domains of the antibody or its fragments. Preferably, the first binding domain is characterized by the presence of three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VH region). The second binding domain preferably also includes the minimum structural requirements of the antibody that enable target binding. More preferably, the second binding domain comprises at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VH region). The first and / or second binding domains are expected to be prepared or obtained by phage display or library screening, rather than by grafting onto a scaffold of CDR sequences derived from existing (monoclonal) antibodies.

[0098] According to the present invention, the binding domain is in the form of one or more polypeptides. Such polypeptides may comprise a proteinaceous portion and a non-proteinaceous portion (e.g., a chemical linker or a chemical crosslinking agent such as glutaraldehyde). Proteins (including their fragments, preferably biologically active fragments, and peptides having typically fewer than 30 amino acids) comprise two or more amino acids linked to each other via covalent peptide bonds (resulting in a chain of amino acids).

[0099] As used herein, the term “polypeptide” typically refers to a group of molecules consisting of more than 30 amino acids. Polypeptides can further form polymers such as dimers, trimers, and higher-order oligomers, i.e., polymers consisting of two or more polypeptide molecules. The polypeptide molecules forming such dimers, trimers, etc., may or may not be identical. The corresponding higher-order structures of such polymers are therefore referred to as homo- or heterodimers, homo- or heterotrimers, and so on. An example of a heteropolymer is an antibody molecule in its natural form, consisting of two identical polypeptide light chains and two identical polypeptide heavy chains. The terms “peptide,” “polypeptide,” and “protein” also refer to naturally modified peptides / polypeptides / proteins that have been modified by post-translational modifications such as glycosylation, acetylation, and phosphorylation. As used herein, “peptide,” “polypeptide,” or “protein” may also be chemically modified, such as pegylation. Such modifications are well known in the art and are described below herein.

[0100] Preferably, the binding domain that binds to the surface antigen of the target cell and / or the binding domain that binds to CD3ε is a human binding domain. Antibodies and antigen-binding polypeptides containing at least one human binding domain avoid some of the problems associated with antibodies or antigen-binding polypeptides having non-human variable and / or constant regions, such as those from rodents (e.g., mice, rats, hamsters, or rabbits). The presence of such rodent-derived proteins can lead to rapid clearance of the antibody or antigen-binding polypeptide or to trigger an immune response by the patient to the antibody or antigen-binding polypeptide. To avoid the use of rodent-derived antibodies or antigen-binding polypeptides, human or fully human antibody / antigen-binding polypeptides can be generated by introducing human antibody function into rodents so that the rodents produce fully human antibodies.

[0101] The ability of YAC to clone and reconstruct megabase-sized human loci and introduce them into mouse germline cells provides a powerful method for elucidating the functional elements of very large or coarsely mapped loci and for generating useful models of human diseases. Furthermore, using such techniques to replace mouse loci with their human equivalents could provide unique insights into the expression and regulation of human gene products during development, their transmission to other systems, and their involvement in disease induction and progression.

[0102] A key practical application of such strategies is the "humanization" of the mouse humoral immune system. Introducing human Ig loci into mice with inactivated endogenous immunoglobulin (Ig) genes provides an opportunity to study the mechanisms underlying programmed antibody expression and construction, as well as their roles in B cell development. Furthermore, such strategies could provide an ideal source for producing fully human monoclonal antibodies (mAbs), a crucial milestone in realizing the potential of antibody therapy in human diseases. Fully human antibodies or antigen-binding polypeptides are expected to minimize the immunogenicity and allergic reactions inherent in mouse mAbs or mouse-derived mAbs, thereby increasing the efficacy and safety of the administered antibody / antigen-binding polypeptide. The use of fully human antibodies or antigen-binding polypeptides is expected to offer significant advantages in the treatment of chronic and recurrent human diseases requiring repeated administration of compounds, such as inflammation, autoimmunity, and cancer.

[0103] One approach to achieving this goal involves manipulating mouse strains lacking mouse antibody production using large fragments of the human Ig locus. This is based on the prediction that such mice would produce a broad repertoire of human antibodies without producing mouse antibodies. Large human Ig fragments are thought to retain broad diversity of variable genes as well as appropriate regulation of antibody production and expression. By utilizing mouse mechanisms for antibody diversification and selection, and for the lack of immune tolerance to human proteins, the human antibody repertoire reproduced in these mouse strains should produce high-affinity antibodies against any target antigen, including human antigens. Using hybridoma technology, antigen-specific human mAbs with the desired specificity can be easily generated and selected. This general strategy was demonstrated in connection with the generation of the first XenoMouse mouse strain (see Green et al. Nature Genetics 7:13-21 (1994)). This XenoMouse strain was engineered using yeast artificial chromosomes (YACs) containing germline-arranged fragments of 245kb and 190kb sizes, respectively, of the human heavy chain locus and kappa light chain locus, containing core sequences of the variable and constant regions. These human Ig-containing YACs proved compatible with the mouse strain for both antibody rearrangement and expression, and were capable of replacing 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 antigen-specific human mAbs. These results also suggested that the introduction of a large portion of the human Ig locus, containing numerous V genes, additional regulatory elements, and the human Ig constant region, could reproduce a substantially complete repertoire characterized by the human humoral response to infection and immunization. More recently, building upon the work of Green et al., the introduction of megabase-sized germline-arranged YAC fragments of the human heavy chain locus and kappa light chain locus resulted in the introduction of over 80% of the human antibody repertoire.See Mendez et al. Nature Genetics 15:146-156 (1997) and U.S. Patent Application No. 08 / 759,620.

[0104] The creation of the XenoMouse mouse is further supported by U.S. Patent Applications No. 07 / 466,008, No. 07 / 610,515, No. 07 / 919,297, No. 07 / 922,649, No. 08 / 031,801, No. 08 / 112,848, No. 08 / 234,145, No. 08 / 376,279, No. 08 / 430,938, No. 08 / 464,584, No. 08 / 464,582, No. 08 / 463,191, No. 08 / 462,837, and No. 0 This is discussed and detailed in Patent Nos. 8 / 486,853, 08 / 486,857, 08 / 486,859, 08 / 462,513, 08 / 724,752 and 08 / 759,620; and in U.S. Patent Nos. 6,162,963, 6,150,584, 6,114,598, 6,075,181 and 5,939,598; and in Japanese Patent Publication 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), European Patent No. 0463151B1, International Publication No. 94 / 02602, International Publication No. 96 / 34096, International Publication No. 98 / 24893, International Publication No. 00 / 76310 and International Publication No. 03 / 47336.

[0105] Another approach, employed by companies including GenPharm International, Inc., utilizes the "mini-locus" method. In this mini-locus method, an exogenous Ig locus is mimicked by including fragments (individual genes) derived from this Ig locus. Thus, one or more VH genes, one or more DH genes, one or more JH genes, a muon constant region, and a second constant region (preferably a gamma constant region) form a construct that is inserted into the animal. This method is based on U.S. Patent No. 5,545,807 to Surani et al., and U.S. Patents No. 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 to Lonberg and Kay, respectively. The document; U.S. Patent No. 5,874,299; and U.S. Patent No. 6,255,458, U.S. Patent No. 5,591,669 and U.S. Patent No. 6,023,010 against Krimpenfort and Berns, U.S. Patent No. 5,612,205; U.S. Patent No. 5,721,367; and U.S. Patent No. 5,789,215 against Berns et al., and U.S. Patent No. 5,643,763 against Choi and Dunn, and GenPharm This is described in U.S. Patent Applications No. 07 / 574,748, No. 07 / 575,962, No. 07 / 810,279, No. 07 / 853,408, No. 07 / 904,068, No. 07 / 990,860, No. 08 / 053,131, No. 08 / 096,762, No. 08 / 155,301, No. 08 / 161,739, No. 08 / 165,699, and No. 08 / 209,741 of International.See also European Patent No. 0546073B1, International Publication No. 92 / 03918, International Publication No. 92 / 22645, International Publication No. 92 / 22647, International Publication No. 92 / 22670, International Publication No. 93 / 12227, International Publication No. 94 / 00569, International Publication No. 94 / 25585, International Publication No. 96 / 14436, International Publication No. 97 / 13852 and International Publication No. 98 / 24884, and U.S. Patent No. 5,981,175. Furthermore, see Taylor et al. (1992), Chen et al. (1993), Tuaillon et al. (1993), Choi et al. (1993), Lonberg et al. (1994), Taylor et al. (1994), Tuaillon et al. (1995), and Fishwild et al. (1996).

[0106] Kirin also demonstrated the production of human antibodies from mice in which large chromosome fragments or entire chromosomes were introduced by microcell fusion. See European Patent Applications No. 773288 and No. 843961. Xenerex Biosciences is developing a potential human antibody production technology in which SCID mice are reconstituted with human lymphocytes, such as B cells and / or T cells. The mice are then immunized with an antigen, which can induce an immune response. See U.S. Patents No. 5,476,996; No. 5,698,767; and No. 5,958,765.

[0107] Human-anti-mouse antibody (HAMA) reactions have led the industry to produce chimeric or otherwise humanized antibodies. However, certain human-chimeric antibody (HACA) reactions are expected to be observed, particularly in chronic or multi-dose antibody use. Therefore, to eliminate concerns and / or effects of HAMA or HACA reactions, it is desirable to provide antigen-binding polypeptides containing human-binding domains to target cell surface antigens and human-binding domains to CD3ε.

[0108] The terms "(specifically) bind," "(specifically) recognize," "(specifically) induce," and "(specifically) react" mean, according to the present invention, that the binding domain interacts with or specifically interacts with a target molecule (antigen), in this specification, with a surface antigen of a target cell and a given epitope or target site on CD3ε, respectively.

[0109] The term "epitope" refers to a site on an antigen to which the binding domain of 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 may also be referred to herein as an "antigenic structure" or "antigenic determinant." Thus, the binding domain is an "antigen interaction site." The binding / interaction is also understood to define "specific recognition."

[0110] An "epitope" can be formed by both consecutive amino acids or discontinuous amino acids that are parallel due to the three-dimensional folding of a protein. A "linear epitope" is an epitope that contains an epitope in which the primary amino acid sequence is recognized. Linear epitopes typically contain at least three or at least four, more commonly at least five, at least six, or at least seven, for example, about eight to about ten amino acids within a specific sequence.

[0111] A "structural epitope," in contrast to a linear epitope, is an epitope in which the primary sequence of amino acids containing the epitope is not the sole element defining the recognized epitope (for example, an epitope in which the primary sequence of amino acids is not necessarily recognized by the binding domain). Generally, structural epitopes contain more amino acids than linear epitopes. In relation to the recognition of structural epitopes, the binding domain recognizes the three-dimensional structure of an antigen, preferably a peptide or protein, or a fragment thereof (in the context of this 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, specific amino acids and / or polypeptide backbone that form the structural epitope are arranged in parallel, thereby enabling the antibody to recognize that epitope. Methods for determining the three-dimensional structure of an epitope include, but are not limited to, X-ray crystallography, two-dimensional nuclear magnetic resonance (2D-NMR) spectroscopy, and site-directed spin labeling and electron paramagnetic resonance (EPR) spectroscopy.

[0112] The method of epitope mapping is described below. When a region (a sequence of adjacent amino acids) of a human target cell's surface antigen protein is exchanged / substituted with the corresponding region of a non-human and non-primate target cell's surface antigen (for example, a mouse target cell's surface antigen, but others such as chicken, rat, hamster, and rabbit are also possible), a reduction in the binding affinity of the binding domain is expected to occur, unless the binding domain is cross-reactive to the non-human and non-primate target cell's surface antigen being used. The aforementioned 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 corresponding region within the human target cell's surface antigen protein, with binding to the corresponding region of the human target cell's surface antigen protein being 100%. The above human target cell surface antigen / non-human target cell surface antigen chimeras are expected to be expressed in CHO cells. Furthermore, it is conceivable that chimeric forms of surface antigens from human target cells and non-human target cells may fuse with the transmembrane and / or cytoplasmic domains of different membrane-bound proteins, such as EpCAM.

[0113] As an alternative or additional method to epitope mapping, several truncated forms of the extracellular domain of human target cell surface antigens may be created to determine specific regions recognized by the binding domain. In these truncated forms, different extracellular target cell surface antigen domains / subdomains or regions are deleted stepwise, starting from the N-terminus. It is hypothesized that truncated target cell surface antigens may be expressed in CHO cells. It is also hypothesized that truncated target cell surface antigens may be fused with the transmembrane and / or cytoplasmic domains of different membrane-bound proteins, such as EpCAM. Furthermore, it is hypothesized that truncated target cell surface antigens may contain a signal peptide domain at their N-terminus, e.g., a signal peptide derived from mouse IgG heavy chain signal peptide. In addition, it is hypothesized that truncated target cell surface antigens may contain a v5 domain at the N-terminus (following the signal peptide) to confirm their precise expression on the cell surface. In truncated target cell surface antigens that no longer contain the target cell surface antigen region recognized by the binding domain, a reduction or loss of binding is expected. The reduction in binding 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 further 100%, when the binding to the entire surface antigen protein (or its extracellular region or domain) of the human target cell is set to 100%.

[0114] A further method for determining the contribution of specific residues of a surface antigen of a target cell to recognition by an antigen-binding polypeptide or binding domain is alanine scanning, which involves substituting each residue to be analyzed with alanine, for example, by site-directed mutagenesis (see, for example, Morrison KL & Weiss GA. Cur Opin Chem Biol. 2001 Jun;5(3):302-7). Alanine is used because, despite mimicking the secondary structure criteria of many other amino acids, it is not bulky and has a chemically inert methyl functional group. If it is desirable to conserve the size of the residue being mutated, sometimes bulky amino acids such as valine or leucine may be used. Alanine scanning is a well-established technique that has been used for a long time.

[0115] The interaction between the binding domain and an epitope or region containing an epitope means that the binding domain exhibits a measurable affinity for an epitope / region containing an epitope on a specific protein or antigen (herein, the surface antigen of a target cell and CD3, respectively), and generally does not exhibit significant reactivity with proteins or antigens other than the surface antigen of the target cell or CD3. "Measurable affinity" includes bindings having an 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 -9If M is present, the binding is considered specific. Whether the binding domain specifically reacts to or binds to a target can be easily tested, in particular, by comparing the reaction of the binding domain to the target protein or antigen with the reaction of the binding domain to the surface antigen of the target cell or to proteins or antigens other than CD3. Preferably, the binding domain of the present invention does not essentially or substantially bind to the surface antigen of the target cell or to proteins or antigens other than CD3 (i.e., the first binding domain cannot bind to proteins other than the surface antigen of the target cell, and the second binding domain cannot bind to proteins other than CD3). Having superior affinity properties compared to other HLE forms is a presumed feature of the antigen-binding polypeptide according to the present invention. Such superior affinity consequently suggests an extended half-life in vivo. The longer half-life of the antigen-binding polypeptide according to the present invention may reduce the duration and frequency of administration, which usually contributes to improved patient compliance. This is particularly important because the antigen-binding polypeptide of the present invention is particularly beneficial to cancer patients who are severely debilitated or even more multi-disease patients.

[0116] The terms "essentially / substantially non-binding" or "unable to bind" mean that the binding domain of the present invention does not bind to the surface antigen of the target cell or any protein or antigen other than CD3, i.e., if binding to the surface antigen of the target cell or CD3 is taken as 100%, it does not show reactivity to any protein or antigen other than the surface antigen of the target cell or CD3 by more than 30%, preferably more than 20%, more preferably more than 10%, and particularly preferably more than 9%, 8%, 7%, 6%, or 5%.

[0117] Specific binding is thought to be brought about by specific motifs within the binding domain and the amino acid sequence of the antigen. Therefore, binding occurs as a result of their primary, secondary, and / or tertiary structures, as well as as a result of secondary modifications of said structures. Specific interaction between the antigen interaction site and its specific antigen can lead to simple binding of the site to the antigen. Furthermore, specific interaction between the antigen interaction site and its specific antigen can alternatively or additionally lead to signal initiation, for example, by inducing conformational changes in the antigen or oligomerization of the antigen.

[0118] The term "variable" refers to a portion of an antibody or immunoglobulin domain (i.e., a "variable domain") that exhibits variability within a sequence and is involved in determining the specificity and binding affinity of a particular antibody. A pair of variable heavy chains (VH) and variable light chains (VL) together form a single antigen-binding site.

[0119] The variability is not uniformly distributed throughout the antibody's variable domain, but rather concentrated in the respective subdomains 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) portions of the variable domain are called "framework" regions (FRMs or FRs), which provide a scaffold for the six CDRs in the three-dimensional space that forms the antigen-binding surface. The naturally occurring heavy and light chain variable domains each contain four FRM regions (FR1, FR2, FR3, and FR4), mostly in 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 collectively in close proximity by the FRMs and, together with the hypervariable regions of other chains, contribute to the formation of the antigen-binding site (see Kabat et al. cited above).

[0120] The term "CDR" and its plural form "CDRs" refer to 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 which constitute the binding properties of the heavy chain variable region (CDR-H1, CDR-H2, and CDR-H3). CDRs contain most of the residues responsible for the specific interaction between antibodies and antigens, and therefore contribute to the functional activity of antibody molecules. CDRs are the primary determinants of antigen specificity.

[0121] The precise definition of CDR boundaries and length follows various classification and numbering schemes. Therefore, CDRs may be represented by Kabat, Chothia, contact, or any other boundary definition, including the numbering schemes described herein. Even with different boundaries, each of these schemes has some overlap in the portions constituting the so-called "hypervariable regions" within the variable sequence. Therefore, the definitions of CDRs by these schemes may differ in length and boundary regions with respect to adjacent framework regions. See, for example, Kabat (a method based on interspecies sequence variability), Chothia (a method based on crystallographic studies of antigen-antibody complexes), and / or MacCallum (Kabat et al., op. cit.; Chothia et al., J.Mol.Biol, 1987, 196:901-917; and MacCallum et al., J.Mol.Biol, 1996, 262:732). Another criterion for characterizing antigen-binding sites 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). Two residue identification techniques can be combined to define hybrid CDRs, insofar as they define overlapping but not identical regions. However, numbering following the so-called Kabat system is preferred.

[0122] Typically, CDRs form loop structures that can be classified as canonical structures. The term "canonical structure" refers to the three-dimensional structure of the main chain adopted by the antigen-binding (CDR) loop. Comparative structural studies have revealed that five of the six antigen-binding loops have only a limited repertoire of available three-dimensional structures. Each canonical structure can be characterized by the twist angle of the polypeptide backbone. Therefore, corresponding loops between antibodies can have very similar three-dimensional structures, despite high amino acid sequence variability being observed in most 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 correlation between the adopted loop structure and the surrounding amino acid sequence. The three-dimensional structure 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 and within the conserved framework (i.e., outside the loop). Therefore, assignment to a particular canonical class can be based on the presence of these key amino acid residues.

[0123] The term "canonical structure" may also include considerations of the linear sequence of an antibody, as classified, for example, by Kabat (Kabat et al., cited above). Kabat's numbering scheme is a widely adopted standard for numbering amino acid residues of antibody variable domains in a consistent manner and is the preferred scheme applied in this invention, as mentioned elsewhere in this specification. Further structural considerations may also be used to determine the canonical structure of an antibody. For example, differences not adequately reflected by Kabat's numbering scheme can be described by the numbering scheme of Chothia et al. and / or revealed by other techniques, such as crystallography and two- or three-dimensional computer modeling. Thus, a given antibody sequence may be classified into a canonical class, among others, from which the appropriate chassis sequence can be identified (for example, based on the requirement to include various canonical structures in a library). The Kabat numbering scheme for antibody amino acid sequences, the structural considerations described by Chothia et al. (cited above), and their significance for interpreting the canonical aspects of antibody structure are described in the literature. The subunit structures and three-dimensional arrangements of various classes of immunoglobulins are well known in the art. For an overview of antibody structures, see Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory, eds. Harlow et al., 1988.

[0124] Light chain CDR3, and especially heavy chain CDR3, can constitute the most important determinants in antigen binding within the light and heavy chain variable regions. In some antigen-binding polypeptides, heavy chain CDR3 appears to constitute the primary contact region between the antigen and the antibody. In vitro selection schemes that alter only the 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 two amino acid residues or more than 26 amino acids.

[0125] In classical full-length antibodies or immunoglobulins, each light (L) chain is linked to the heavy (H) chain by a single 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") domain does not directly participate in antigen binding but exhibits various effector functions, such as antibody-dependent, cell-mediated cytotoxicity, and complement activation. The Fc region of the antibody is contained within the heavy chain constant domain and can interact with Fc receptors located, for example, on the cell surface.

[0126] The sequences of antibody genes after construction and somatic mutation are extremely diverse, and these diversified genes are 10 10 It is presumed to encode different 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 that is derived in whole or in part from at least one sequence encoding at least one immunoglobulin. This 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, this sequence may be generated from cells in response to a rearrangement, for example, an in vitro stimulus. Alternatively, some or all of this sequence may be obtained by DNA splicing, nucleotide synthesis, mutagenesis, and other methods (see, for example, U.S. Patent No. 5,565,332). A repertoire may consist of only one sequence or may consist of multiple sequences, including those in a genetically diverse collection.

[0127] In relation to the present invention, the term "Fc moiety" or "Fc monomer" means a polypeptide comprising at least one domain having the function of the CH2 domain and at least one domain having the function of the CH3 domain of an immunoglobulin molecule. As is clear from the term "Fc monomer," polypeptides containing these CH domains are "polypeptide monomers." An Fc monomer may be a polypeptide comprising at least a fragment of the constant region of an immunoglobulin, excluding the first constant region immunoglobulin domain (CH1) of the heavy chain, but maintaining the functional portions of at least one CH2 domain and one CH3 domain (where the CH2 domain is on the amino-terminal side of the CH3 domain). In a preferred embodiment of this definition, an Fc monomer may be a polypeptide constant region comprising a portion of the Ig-Fc hinge region, a CH2 region and a CH3 region (where the hinge region is on the amino-terminal side of the CH2 domain). The hinge region of the present invention is intended to promote dimerization. Such Fc polypeptide molecules can be obtained, for example, by papain digestion of an immunoglobulin region (which naturally produces a dimer of two Fc polypeptides), but are not limited thereto. In another embodiment of this definition, the Fc monomer may be a polypeptide region containing parts of the CH2 and CH3 regions. Such Fc polypeptide molecules can be obtained, for example, by pepsin digestion of an immunoglobulin molecule, but are not limited thereto. In one embodiment, the polypeptide sequence of the Fc monomer is substantially similar to the Fc polypeptide sequences of the IgG1Fc, IgG2Fc, IgG3Fc, IgG4Fc, IgM Fc, IgA Fc, IgD Fc, and IgE Fc regions. (See, for example, Padlan, Molecular Immunology, 31(3), 169-217 (1993)). Because several variants exist among immunoglobulins, and simply for clarity, the Fc monomer refers to the two heavy chain constant-region immunoglobulin domains at the end of IgA, IgD, and IgG, and the three heavy chain constant-region immunoglobulin domains at the end of IgE and IgM. As described above, the Fc monomer may also contain a flexible hinge on the N-terminal side of these domains.In the case of IgA and IgM, the Fc monomer may contain a J chain. In the case of IgG, the Fc portion contains immunoglobulin domains CH2 and CH3, as well as a hinge between the first two domains and CH2. Although the boundaries of the Fc portion may differ, an example of a human IgG heavy chain Fc portion containing a functional hinge, CH2 and CH3 domains can be defined, for example, by Kabat numbering, as containing residues D231 (of the hinge domain - corresponding to D234 in Table 1 below) to P476 or L476 (in the case of IgG4) at the carboxyl end of the CH3 domain, respectively. Two Fc portions or Fc monomers fused to each other via a peptide linker define a third domain of the antigen-binding polypeptide of the present invention, which can also be defined as an scFc domain.

[0128] In one embodiment of the present invention, the scFc domain disclosed herein, or the Fc monomers fused together, are assumed to be contained only in the third domain of the antigen-binding polypeptide.

[0129] In accordance with the present invention, the IgG hinge region can be identified by similarity using Kabat numbering as described in Table 1. In line with the above, the hinge domain / region of the present invention is assumed to contain amino acid residues corresponding to a sequence of IgG1 sequences from D234 to P243 by Kabat numbering. Similarly, the hinge domain / region of the present invention is assumed to contain or consist of the IgG1 hinge sequence DKTHTCPPCP (SEQ ID NO: 182) (corresponding to the D234-P243 stretch shown in Table 1 below - variants of the sequence are also assumed, 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 polypeptide is removed by an N314X substitution (where X is any amino acid other than Q). The substitution is preferably an N314G substitution. In a more preferred embodiment, the 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).

[0130] The third domain of the antigen-binding polypeptide of the present invention may also comprise or consist of DKTHTCPPCP(SEQ ID NO: 182)(i.e., hinge)-CH2-CH3-linker-DKTHTCPPCP(SEQ ID NO: 182)(i.e., hinge)-CH2-CH3 in the order of amino to carboxyl. The peptide linker of the aforementioned antigen-binding polypeptide is characterized in a preferred embodiment by the amino acid sequence Gly-Gly-Gly-Gly-Ser, i.e., Gly4Ser(SEQ ID NO: 187), or a polymer thereof, i.e., (Gly4Ser)x, where x is an integer of 5 or more (e.g., 5, 6, 7, 8, etc., or more), with 6 being preferred ((Gly4Ser)6). The construct may further comprise the aforementioned substitution N314X, preferably N314G, and / or further substitutions V321C and R309C. In preferred embodiments of the antigen-binding polypeptide of the present invention as defined above herein, the second domain is envisioned to bind to an extracellular epitope of human and / or macaque CD3ε chain.

[0131] [Table 1]

[0132] In further embodiments of the present invention, the hinge domain / region comprises or consists of the IgG2 subtype hinge sequence ERKCCVECPPCP (SEQ ID NO: 183), the IgG3 subtype hinge sequence ELKTPLDTTHTCPRCP (SEQ ID NO: 184) or ELKTPLGDTTHTCPRCP (SEQ ID NO: 185), and / or the IgG4 subtype hinge sequence ESKYGPPCPSCP (SEQ ID NO: 186). The IgG1 subtype hinge sequence may be the following sequence EPKSCDKTHTCPPCP (as shown in Table 1 and SEQ ID NO: 183). Thus, these core hinge regions are also conceivable in connection with the present invention.

[0133] The locations and sequences of the IgG CH2 and IgG CH3 domains can be identified by similarity using Kabat numbering, as described in Table 2.

[0134] [Table 2]

[0135] In one embodiment of the present invention, the amino acid residues highlighted in bold within the CH3 domain of the first or both Fc monomers are deleted.

[0136] The peptide linker that fuses the polypeptide monomers of the third domain ("Fc moiety" or "Fc monomer") to each other preferably comprises at least 25 amino acid residues (e.g., 25, 26, 27, 28, 29, 30). More preferably, the peptide linker comprises at least 30 amino acid residues (e.g., 30, 31, 32, 33, 34, 35). It is also preferable that the linker comprises up to 40 amino acid residues, more preferably up to 35 amino acid residues, and most preferably just 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: 187), or a polymer thereof, i.e., (Gly4Ser)x, where x is an integer of 5 or greater (e.g., 6, 7, or 8). Preferably, the integer is 6 or 7, and more preferably, the integer is 6.

[0137] When a linker is used to fuse a first domain with a second domain, or to fuse a first or second domain with a third domain, the linker is preferably of sufficient length and sequence to ensure that each of the first and second domains independently maintains their different binding specificities. In the case of peptide linkers connecting at least two binding domains (or two variable domains) within the antigen-binding polypeptide of the present invention, the peptide linkers are preferably those containing only a few amino acid residues, for example, 12 or fewer amino acid residues. Therefore, peptide linkers with 12, 11, 10, 9, 8, 7, 6, or 5 amino acid residues are preferred. Of the assumed peptide linkers having fewer than 5 amino acids, linkers containing 4, 3, 2, or 1 amino acid and rich in Gly are preferred. A preferred embodiment of the peptide linker for fusion of the first and second domains is shown in SEQ ID NO: 1. A preferred linker embodiment of the peptide linker for fusion of the second and third domains is a (Gly)4-linker, or G4-linker.

[0138] In relation to one of the above-mentioned "peptide linkers," a particularly preferred "single" amino acid is Gly. Thus, the aforementioned peptide linker may consist of a single amino acid Gly. In preferred embodiments of the present invention, the peptide linker is characterized by the amino acid sequence Gly-Gly-Gly-Gly-Ser, i.e., Gly4Ser (SEQ ID NO: 187), or a polymer thereof, i.e., (Gly4Ser)x, where x is an integer of 1 or more (e.g., 2 or 3). Preferred linkers are shown in SEQ ID NOs: 1 to 12. Features of the peptide linkers, including not promoting secondary structure, 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). Furthermore, a peptide linker that does not promote any secondary structure is preferred. The linking 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-stranded constructs and expressing them in mammalian cells or bacteria are well known in the art (e.g., International Publication No. 99 / 54440 or Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 2001).

[0139] In antigen-binding polypeptides or preferred embodiments of the present invention, the first and second domains form an antigen-binding polypeptide in a form selected from the group consisting of (scFv)2, scFv-single-domain mAb, diabody, and any oligomer of these forms.

[0140] In particular preferred embodiments, and as described in the accompanying examples, the first and second domains of the antigen-binding polypeptide of the present invention are a "bispecific single-strand antigen-binding polypeptide," more preferably a bispecific "single-strand Fv" (scFv). The two domains of the Fv fragment, VL and VH, are encoded by separate genes, but they can be linked by a synthetic linker as described above herein, which allows them to be prepared using recombination as a single protein chain paired so that the VL and VH regions form a monovalent molecule; see, for example, Huston et al. (1988) Proc. Natl. Acad. Sci USA 85:5879-5883. These antibody fragments can be obtained using the prior art known to those skilled in the art, and their function can be evaluated in the same manner as that of a complete or full-length antibody. Therefore, a single-chain variable fragment (scFv) is a fusion protein of the variable region (VH) of the heavy chain and the variable region (VL) of the light chain of an immunoglobulin, typically linked by a short linker peptide of about 10 to 25 amino acids, preferably about 15 to 20 amino acids. The linker is usually rich in glycine for flexibility and serine or threonine for solubility, and may connect the N-terminus of VH to the C-terminus of VL or vice versa. This protein retains the specificity of the original immunoglobulin despite the removal of the constant region and the introduction of a linker.

[0141] Bispecific single-chain antigen-binding polypeptides are known in the art and are described in International Publication No. 99 / 54440, Mack, J. (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, and Kipriyanov, J.Mol.Biol., (1999), 293, 41-56. The techniques described for the production of single-chain antibodies (see, in particular, U.S. Patent No. 4,946,778, Kontermann and Duebel (2010), cited above, and Little (2009), cited above) may be adapted for the production of single-chain antigen-binding polypeptides that specifically recognize selected targets.

[0142] Bivalent (also called divalent) or bispecific single-strand variable fragments (bi-scFv or di-scFv having the form (scFv)2) can be produced by linking two scFv molecules (for example, using the linkers described above 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 2 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 constructing a single peptide chain having two VH regions and two VL regions to obtain a tandem scFv (see, for example, Kufer P. et al., (2004) Trends in Biotechnology 22(5):238-244). Another possibility is to construct the scFv molecule using a linker peptide that is too short (e.g., about 5 amino acids) to fold the two variable regions together, and then dimerize the scFv. This type is known as a diabody (see, for example, Hollinger, Philipp et al., (July 1993) Proceedings of the National Academy of Sciences of the United States of America 90(14):6444-8).

[0143] In accordance with the present invention, the first domain, the second domain, or the first and second domains may comprise a single-domain antibody, a variable domain, or at least the CDR of a single-domain antibody. A single-domain antibody comprises only one (monomer) antibody variable domain that can selectively bind to a specific antigen independently of other V regions or domains. The initial single-domain antibodies were produced from heavy-chain antibodies found in camels, and these are V H It is called an H fragment. Cartilaginous fish also have V NARIt possesses heavy chain antibodies (IgNARs) from which single-domain antibodies called fragments can be obtained. An alternative method is to split a dimeric variable domain derived from a common immunoglobulin, for example, from human or rodent, into monomers, thereby obtaining VH or VL as a single-domain Ab. Most research on single-domain antibodies is currently based on heavy chain variable domains, but it has also been shown that light chain-derived nanobodies can specifically bind to target epitopes. Examples of single-domain antibodies are called sdAbs, nanobodies, or single-variable-domain antibodies.

[0144] Therefore, (single-domain mAb)2 is V H , V L , V H H and V NAR It is a monoclonal antigen-binding polypeptide composed of at least two single-domain monoclonal antibodies individually selected from the group including the above. The linker is preferably in the form of a peptide linker. Similarly, "scFv-single-domain mAb" is a monoclonal antigen-binding polypeptide composed of at least one of the above single-domain antibodies and one of the above scFv molecules. In this case as well, the linker is preferably in the form of a peptide linker.

[0145] Whether an antigen-binding polypeptide competes to bind to another given antigen-binding polypeptide 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. Similar to avidin-coated ELISA plates, each of these beads can be used as a substrate when reacting with biotinylated proteins, and the assay can be performed on it. The antigen is coated onto the beads, then pre-coated with a first antibody. A second antibody is added to confirm any further binding. Flow cytometry is a possible method for reading the results.

[0146] T cells, or T lymphocytes, are a type of lymphocyte (a type of white blood cell) that plays a central role in cellular immunity. Several subsets of T cells exist, each with different functions. T cells can be distinguished from other lymphocytes, such as B cells and NK cells, by the presence of a T cell receptor (TCR) on their cell surface. The TCR is involved in the recognition of antigens bound to major histocompatibility complex (MHC) molecules and is composed of two distinct 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 related enzymes, co-receptors, specialized adapter molecules, and activated or released transcription factors.

[0147] The CD3 receptor complex is a protein complex composed of four chains. In mammals, this complex contains a CD3γ (gamma) chain, a 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 cell surface proteins of the highly related immunoglobulin superfamily, each containing a single extracellular immunoglobulin domain. The intracellular tail of the CD3 molecule contains a single conserved motif, known as the immunoreceptor activation tyrosine motif, or ITAM, which is essential for TCR signaling. The CD3 epsilon molecule is a polypeptide encoded by the CD3E gene located on human chromosome 11. The most preferred CD3 epsilon epitopes are those in the range of amino acid residues 1-27 of the human CD3 epsilon extracellular domain. The antigen-binding polypeptide according to the present invention is expected to exhibit less undesirable nonspecific T cell activity in certain immunotherapies, which is typically and advantageously limited. In other words, this reduces the risk of side effects.

[0148] Lysis of redirected target cells via T cell recruitment by multispecific, or at least bispecific, antigen-binding polypeptides is accompanied by the formation of cytolytic synapses and the delivery of perforin and granzyme. The bound T cells are capable of sequential target cell lysis and are unaffected by immune evasion mechanisms that prevent peptide antigen processing and presentation or clonal T cell differentiation; see, for example, International Publication No. 2007 / 042261.

[0149] The cytotoxicity mediated by the antigen-binding polypeptide of the present invention can be measured by various methods. Effector cells may 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 the surface antigen of the macaque target cell bound by the first domain, the effector cells should also be of macaque origin, such as a macaque T cell line, e.g., 4119LnPx. The target cells should express the surface antigen of the target cell, e.g., the surface antigen of a human or macaque target cell (at least its extracellular domain). The target cells may be a cell line (e.g., CHO) that is stably or transiently transfected with the surface antigen of the target cell, e.g., the surface antigen of a human or macaque target cell. Alternatively, the target cells may be a native target cell line that expresses the surface antigen of the target cell. Typically, EC 50 The value is expected to be lower in target cell lines that express high levels of the target cell's surface antigen on the cell surface. The effector-to-target cell (E:T) ratio is usually about 10:1, but this can vary. The cytotoxic activity of the target cell's surface antigen × CD3 bispecific antigen-binding polypeptide is 51It can be measured in a Cr release assay (incubation time of approximately 18 hours) or a cytotoxicity assay using FACS (incubation time of approximately 48 hours). The incubation time (cytotoxicity response) of the assay can also be changed. Other methods for measuring cytotoxicity are well known to those skilled in the art and include MTT or MTS assays, ATP system assays including bioluminescence assays, sulforhodamine B (SRB) assays, WST assays, clonality assays, and ECIS technology.

[0150] The cytotoxic activity mediated by the surface antigen × CD3 bispecific antigen-binding polypeptide of the target cells of the present invention is preferably measured by a cell-based cytotoxicity assay. Cytotoxic activity is also, 51 It can also be measured by a Cr release assay. Cytotoxic activity is EC 50 This is expressed as a value, which corresponds to the half-effect concentration (the concentration of antigen-binding polypeptide that induces an intermediate cytotoxic response between baseline and maximum). Preferably, the EC of the target cell surface antigen × CD3 bispecific antigen-binding polypeptide. 50 The values ​​are ≤5000pM or ≤4000pM, more preferably ≤3000pM or ≤2000pM, even more preferably ≤1000pM or ≤500pM, even more preferably ≤400pM or ≤300pM, even more preferably ≤200pM, even more preferably ≤100pM, even more preferably ≤50pM, even more preferably ≤20pM or ≤10pM, and most preferably ≤5pM.

[0151] In various assays, the given EC above 50 The value can be measured. Stimulated / enriched CD8 + When T cells are used as effector cells, EC is superior to unstimulated PBMCs. 50 Those skilled in the art are aware that the value can be expected to be lower. Furthermore, when target cells express a large number of target cell surface antigens, the EC is lower compared to rats with low target expression. 50 The value can be expected to be lower. For example, stimulated / enriched human CD8 +When using T cells as effector cells (and when using cells transfected with the surface antigen of target cells such as CHO cells or human cell lines positive for the surface antigen of target cells as target cells), the EC of the target cell's surface antigen × CD3 bispecific antigen-binding polypeptide 50 The value is preferably ≤1000pM, more preferably ≤500pM, even more preferably ≤250pM, even more preferably ≤100pM, even more preferably ≤50pM, even more preferably ≤10pM, and most preferably ≤5pM. When human PBMCs are used as effector cells, the EC of the target cell surface antigen × CD3 bispecific antigen-binding polypeptide 50 The value is preferably ≤5000pM or ≤4000pM (especially when the target cells are human cell lines positive for the surface antigen of the target cells), more preferably ≤2000pM (especially when the target cells are cells transfected with the surface antigen of the target cells, such as CHO cells), more preferably ≤1000pM or ≤500pM, even more preferably ≤200pM, even more preferably ≤150pM, even more preferably ≤100pM, and most preferably ≤50pM. When a macaque T cell line such as LnPx4119 is used as an effector cell and a cell line transfected with the surface antigen of macaque target cells such as CHO cells is used as the target cell line, the EC of the target cell surface antigen × CD3 bispecific antigen-binding polypeptide 50 The value is preferably ≤2000pM or ≤1500pM, more preferably ≤1000pM or ≤500pM, even more preferably ≤300pM or ≤250pM, even more preferably ≤100pM, and most preferably ≤50pM.

[0152] Preferably, the target cell surface antigen × CD3 bispecific antigen-binding polypeptide of the present invention does not induce or mediate the lysis of target cell surface antigen-negative cells, such as CHO cells, or substantially does not induce or mediate the lysis. The terms “does not induce lysis,” “substantially does not induce lysis,” “does not mediate lysis,” or “substantially does not mediate lysis” mean that, with the lysis of target cell surface antigen-positive human cell lines being 100%, the antigen-binding polypeptide of the present invention does not induce or mediate the lysis of target cell surface antigen-negative cells by more than 30%, preferably more than 20%, more preferably more than 10%, and particularly preferably more than 9%, 8%, 7%, 6%, or 5%. This is typically true at antigen-binding polypeptide concentrations up to 500 nM. Those skilled in the art know how to measure cell lysis without further effort. Furthermore, specific instructions for measuring cell lysis are taught herein.

[0153] The difference in cytotoxic activity between the monomeric isoform and dimeric isoform of a surface antigen × CD3 bispecific antigen-binding polypeptide of individual target cells is called the "potency gap." This potency gap is, for example, the EC of the monomeric form of the molecule. 50 Value and dimeric form of EC 50 The efficacy gap of the target cell surface antigen × CD3 bispecific antigen-binding polypeptide of the present invention is preferably ≤5, more preferably ≤4, even more preferably ≤3, even more preferably ≤2, and most preferably ≤1.

[0154] The first and / or second (or any further) binding domains of the antigen-binding polypeptide of the present invention are preferably interspecies-specific in members of the order Mammalia of primates. Interspecies-specific CD3-binding domains are described, for example, in International Publication No. 2008 / 119567. According to one embodiment, the first and / or second binding domains bind to the surface antigens / CD3 of target cells of primates, including (but not limited to) New World primates (such as the common marmoset (Callithrix jacchus), cotton-top tamarin (Saguinus Oedipus), or squirrel monkey (Saimiri sciureus)), Old World primates (such as baboons and macaques), gibbons, and non-human homininae, respectively, in addition to binding to the surface antigens / CD3 of human target cells.

[0155] In one embodiment of the antigen-binding polypeptide of the present invention, the first domain binds to a surface antigen of a human target cell and further binds to a surface antigen of a macaque target cell, such as a surface antigen of a cynomolgus monkey (Macaca fascicularis) target cell, more preferably to a surface antigen of a macaque target cell expressed on the surface of a macaque cell. The affinity of the first binding domain for the surface antigen of a macaque target cell 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, most preferably ≤0.05 nM, or even more preferably ≤0.01 nM.

[0156] Preferably, the binding affinity gap of the antigen-binding polypeptide according to the present invention between the surface antigen of macaque target cells and the surface antigen of human target cells [ma target cell surface antigen: hu target cell surface antigen] (determined, for example, by BiaCore or scatchard analysis) is <100, preferably <20, more preferably <15, even more preferably <10, even more preferably <8, even more preferably <6, and most preferably <2. The preferred range for the binding affinity gap of the antigen-binding polypeptide according to the present invention between the surface antigen of macaque target cells and the surface antigen of human target cells 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.

[0157] The second (binding) domain of the antigen-binding polypeptide of the present invention binds to human CD3 epsilon and / or macaque CD3 epsilon. In preferred embodiments, the second domain further binds to CD3 epsilon of common marmoset (Callithrix jacchus), cotton-top tamarin (Saguinus Oedipus), or squirrel monkey (Saimiri sciureus). The common marmoset (Callithrix jacchus) and the cotton-top tamarin (Saguinus Oedipus) are New World primates belonging to the family Callitrichidae, while the squirrel monkey (Saimiri sciureus) is a New World primate belonging to the family Cebidae.

[0158] In the antigen-binding polypeptide of the present invention, the second binding domain that binds to the extracellular epitope of human and / or macaque CD3 preferably includes a VL region comprising CDR-L1, CDR-L2, and CDR-L3 selected from the following: (a) CDR-L1 as shown in Sequence ID 27 of International Publication No. 2008 / 119567, CDR-L2 as shown in Sequence ID 28 of International Publication No. 2008 / 119567, and CDR-L3 as shown in Sequence ID 29 of International Publication No. 2008 / 119567; (b) CDR-L1 as shown in Sequence ID 117 of International Publication No. 2008 / 119567, CDR-L2 as shown in Sequence ID 118 of International Publication No. 2008 / 119567, and CDR-L3 as shown in Sequence ID 119 of International Publication No. 2008 / 119567; and (c) CDR-L1 as shown in Sequence ID 153 of International Publication No. 2008 / 119567, CDR-L2 as shown in Sequence ID 154 of International Publication No. 2008 / 119567, and CDR-L3 as shown in Sequence ID 155 of International Publication No. 2008 / 119567.

[0159] In a more preferred embodiment of the antigen-binding polypeptide of the present invention, the second domain that binds to the extracellular epitope of the human and / or macaca CD3 epsilon chain comprises a VH region including CDR-H1, CDR-H2, and CDR-H3 selected from the following: (a) CDR-H1 as shown in Sequence ID 12 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 13 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 14 of International Publication No. 2008 / 119567; (b) CDR-H1 as shown in Sequence ID 30 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 31 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 32 of International Publication No. 2008 / 119567; (c) CDR-H1 as shown in Sequence ID 48 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 49 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 50 of International Publication No. 2008 / 119567; (d) CDR-H1 as shown in Sequence ID 66 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 67 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 68 of International Publication No. 2008 / 119567; (e) CDR-H1 as shown in Sequence ID 84 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 85 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 86 of International Publication No. 2008 / 119567; (f) CDR-H1 as shown in Sequence ID 102 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 103 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 104 of International Publication No. 2008 / 119567; (g) CDR-H1 as shown in Sequence ID 120 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 121 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 122 of International Publication No. 2008 / 119567; (h) CDR-H1 as shown in Sequence ID 138 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 139 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 140 of International Publication No. 2008 / 119567; (i) CDR-H1 as shown in Sequence ID 156 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 157 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 158 of International Publication No. 2008 / 119567; and (j) CDR-H1 as shown in Sequence ID 174 of International Publication No. 2008 / 119567, CDR-H2 as shown in Sequence ID 175 of International Publication No. 2008 / 119567, and CDR-H3 as shown in Sequence ID 176 of International Publication No. 2008 / 119567.

[0160] In a preferred embodiment of the antigen-binding polypeptide of the present invention, the three groups of VL CDRs described above are combined with the ten groups of VH CDRs described above within a second binding domain to form a (30) group comprising CDR-L1 to 3 and CDR-H1 to 3, respectively.

[0161] In the antigen-binding polypeptide of the present invention, the second domain that binds to CD3 preferably includes a VL region selected from the group consisting of VL regions as 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 International Publication No. 200 / 119567, or as shown in SEQ ID NO. 200.

[0162] The second domain that binds to CD3 may also preferably include a VH region selected from the group consisting of VH regions as shown in Sequence IDs 15, 19, 33, 37, 51, 55, 69, 73, 87, 91, 105, 109, 123, 127, 141, 145, 159, 163, 177, or 181 of International Publication No. 2008 / 119567, or as shown in Sequence ID 201.

[0163] More preferably, the antigen-binding polypeptide of the present invention is characterized by a second domain that binds to CD3, comprising a VL region and a VH region selected from the group consisting of the following: (a) the VL area as shown in Sequence ID No. 17 or 21 of International Publication No. 2008 / 119567 and the VH area as shown in Sequence ID No. 15 or 19 of International Publication No. 2008 / 119567; (b) the VL area as shown in Sequence ID 35 or 39 of International Publication No. 2008 / 119567 and the VH area as shown in Sequence ID 33 or 37 of International Publication No. 2008 / 119567; (c) The VL area as shown in Sequence ID 53 or 57 of International Publication No. 2008 / 119567 and the VH area as shown in Sequence ID 51 or 55 of International Publication No. 2008 / 119567; (d) The VL area as shown in Sequence ID 71 or 75 of International Publication No. 2008 / 119567 and the VH area as shown in Sequence ID 69 or 73 of International Publication No. 2008 / 119567; (e) the VL area as shown in Sequence ID 89 or 93 of International Publication No. 2008 / 119567 and the VH area as shown in Sequence ID 87 or 91 of International Publication No. 2008 / 119567; (f) The VL area as shown in Sequence ID 107 or 111 of International Publication No. 2008 / 119567 and the VH area as shown in Sequence ID 105 or 109 of International Publication No. 2008 / 119567; (g) the VL area as shown in Sequence ID No. 125 or 129 of International Publication No. 2008 / 119567 and the VH area as shown in Sequence ID No. 123 or 127 of International Publication No. 2008 / 119567; (h) The VL area as shown in Sequence ID No. 143 or 147 of International Publication No. 2008 / 119567 and the VH area as shown in Sequence ID No. 141 or 145 of International Publication No. 2008 / 119567; (i) the VL area as shown in Sequence ID No. 161 or 165 of International Publication No. 2008 / 119567 and the VH area as shown in Sequence ID No. 159 or 163 of International Publication No. 2008 / 119567; and (j) VL areas as shown in Sequence ID 179 or 183 of International Publication No. 2008 / 119567 and VH areas as shown in Sequence ID 177 or 181 of International Publication No. 2008 / 119567.

[0164] A second domain that binds to CD3, including the VL region as shown in SEQ ID NO: 200 and the VH region as shown in SEQ ID NO: 201, is also preferred in relation to the antigen-binding polypeptide of the present invention.

[0165] According to a preferred embodiment of the antigen-binding polypeptide of the present invention, the first and / or second domains have the following configuration: the pair of VH and VL regions is in the form of a single-chain antibody (scFv). The VH and VL regions are arranged in the order VH-VL or VL-VH. Preferably, the VH region is located at the N-terminus of the linker sequence and the VL region is located at the C-terminus of the linker sequence.

[0166] Preferred embodiments of the antigen-binding polypeptide described above of the present invention are characterized by a second domain that binds to CD3, 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 in International Publication No. 2008 / 119567, or the amino acid sequence shown in SEQ ID NO: 202.

[0167] Covalent modifications of antigen-binding polypeptides are also included within the scope of the present invention, which, while not necessarily, are generally performed post-translation. For example, some types of covalent modifications of antigen-binding polypeptides are introduced intramolecularly by reacting specific amino acid residues of the antigen-binding polypeptide with an organic derivatizing agent that can react with selected side chains or N-terminal or C-terminal residues.

[0168] Cysteinyl residues most commonly react with α-haloacetates (and corresponding amines), such as chloroacetic acid or chloroacetamide, to produce carboxymethyl or carboxyamidemethyl derivatives. Cysteinyl residues can also be derivatized by reactions with bromotrifluoroacetone, α-bromo-β-(5-imidazoyl)propionic acid, chloroacetyl phosphate, N-alkylmaleimide, 3-nitro-2-pyridyl disulfide, methyl 2-pyridyl disulfide, p-chloromercrinebenzoic acid, 2-chloromercrine-4-nitrophenol, or chloro-7-nitrobenzo-2-oxa-1,3-diazole.

[0169] Histidyl residues are derivatized by reaction with diethyl pyrocarbonate at pH 5.5–7.0 because this agent is relatively specific to the histidyl side chain. Para-bromophenacyl bromide is also useful, and this reaction is preferably carried out in 0.1 M sodium cacodylate at pH 6.0. Ricinyl and amino-terminal residues react with succinic anhydride or other carboxylic acid anhydrides. Derivatization with these agents has the effect of reversing the charge of the ricinyl residue. Other suitable reagents for derivatization of alpha-amino-containing residues include imide esters such as methyl picolinimide; pyridoxal phosphate; pyridoxal; chloroborohydride; trinitrobenzenesulfonic acid; O-methylisourea; 2,4-pentanedione; and transaminase-catalyzed reactions with glyoxylates.

[0170] Arginine residues are modified by reaction with one or more conventional reagents, particularly phenylglyoxal, 2,3-butanedione, 1,2-cyclohexanedione, and ninhydrin. Due to the high pKa of the guanidine functional group, derivatization of arginine residues requires the reaction to be carried out under alkaline conditions. Furthermore, these reagents can react with the lysine group and the arginine epsilon-amino group.

[0171] Specific modification of tyrosyl residues may be performed, particularly to introduce spectral labeling to tyrosyl residues through reactions with aromatic diazonium compounds or tetranitromethane. Most commonly, N-acetylimidazole and tetranitromethane are used to form O-acetyltyrosyl species and 3-nitro derivatives, respectively. 125 I or 131 The chloramine T method described above, which involves iodizing tyrosyl residues using I to prepare a labeled protein for use in radioimmunoassays, is preferred.

[0172] The carboxyl side group (aspartyl or glutamyl) is selectively modified by reaction with a carbodiimide (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, the aspartyl and glutamyl residues are converted to asparaginyl and glutamyl residues by reaction with ammonium ions.

[0173] Derivatization with difunctional substances is useful for crosslinking the antigen-binding polypeptide of the present invention to a water-insoluble support matrix or support surface for use in various methods. Commonly used crosslinking agents include, for example, N-hydroxysuccinimide esters such as 1,1-bis(diazoacetyl)-2-phenylethane, glutaraldehyde, and esters with 4-azidosalicylic acid, homodifunctional imide esters including disuccinimidyl esters such as 3,3'-dithiobis(succinimidylpropionate), and difunctional maleimides such as bis-N-maleimide-1,8-octane. Derivatizing agents such as methyl-3-[(p-azidophenyl)dithio]propioimidate yield photoactivatable intermediates that can form crosslinks in the presence of light. Instead, reactive, water-insoluble matrices such as cyanide-activated carbohydrates, as well as reactive substrates described in U.S. Patents No. 3,969,287; No. 3,691,016; No. 4,195,128; No. 4,247,642; No. 4,229,537; and No. 4,330,440, are used for protein immobilization.

[0174] Other modifications of antigen-binding polypeptides are also considered herein. For example, another type of covalent modification of antigen-binding polypeptides involves linking antigen-binding polypeptides to various non-proteinogenic polymers, including, but not limited to, various polyols in the form described in U.S. Patent Nos. 4,640,835; 4,496,689; 4,301,144; 4,670,417; 4,791,192 or 4,179,337, such as polyethylene glycol, polypropylene glycol, polyoxyalkylene or copolymers of polyethylene glycol and polypropylene glycol. In addition, as is known in the art, amino acid substitutions can be made at various positions within the antigen-binding polypeptide to facilitate the addition of polymers such as PEG.

[0175] Suitable protein-based fluorescent labels also include green fluorescent proteins containing GFP (Chalfie et al., 1994, Science 263:802-805) and EGFP (Clontech Laboratories, Inc., Genbank accession number U55762) from Renilla, Ptilosarcus, or Aequorea species, and blue fluorescent proteins (BFP, Quantum Biotechnologies, Inc. 1801 de Maisonneuve Blvd. West, 8) th 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.) Examples include, but are not limited to, al., 1988, Proc. Natl. Acad. Sci. USA 85:2603-2607, and Renilla (International Publication No. 92 / 15673, International Publication No. 95 / 07463, International Publication No. 98 / 14605, International Publication No. 98 / 26277, International Publication No. 99 / 49019, U.S. Patent 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; and 5,925,558).

[0176] The antigen-binding polypeptides of the present invention may also include additional domains that, for example, aid in the isolation of the molecule or relate to the adaptation of the molecule's pharmacokinetic profile. Domains that aid in the isolation of the antigen-binding polypeptide may be selected from an isolation method, such as a peptide motif that can be captured by an isolation column, or a secondarily introduced portion. Non-limiting embodiments 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 their variants (e.g., StrepII tags), and His tags. All antigen-binding polypeptides disclosed herein, characterized by identified CDRs, may include His-tagged domains, commonly known as repeats 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 located at either the N-terminus or the C-terminus of the antigen-binding polypeptide, but is preferably located at the C-terminus. Most preferably, a hexahistidine tag (HHHHHH) (SEQ ID NO: 199) is attached to the C-terminus of the antigen-binding polypeptide according to the present invention via a peptide bond. In addition, a PLGA-PEG-PLGA conjugate system may be combined with the polyhistidine tag for sustained-release applications and improved pharmacokinetic profiles.

[0177] Amino acid sequence modifications of antigen-binding polypeptides described herein are also intended. For example, improvement of the binding affinity and / or other biological properties of the antigen-binding polypeptide may be desirable. Amino acid sequence variants of antigen-binding polypeptides are prepared by introducing appropriate nucleotide changes into the nucleic acid of the antigen-binding polypeptide or by peptide synthesis. All of the amino acid sequence modifications described below should result in antigen-binding polypeptides that continue to retain the desired biological activity of the unmodified parent molecule (binding to surface antigens and CD3 on target cells).

[0178] The term "amino acid" or "amino acid residue" generally refers to an amino acid with a definition recognized in the art, such as an amino acid 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 (Ile 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), but modified, synthesized, or rare amino acids may be used as needed. Generally, amino acids can be classified by the presence of nonpolar side chains (e.g., Ala, Cys, Ile, Leu, Met, Phe, Pro, Val); negatively charged side chains (e.g., Asp, Glu); positively charged side chains (e.g., Arg, His, Lys); or uncharged polar side chains (e.g., Asn, Cys, Gln, Gly, His, Met, Phe, Ser, Thr, Trp, and Tyr).

[0179] Amino acid modifications include, for example, deletions of residues within the amino acid sequence of an antigen-binding polypeptide, and / or insertions of residues, and / or substitutions of residues. Any combination of deletions, insertions, and substitutions is performed to reach the final polypeptide, provided that the final polypeptide retains the desired characteristics. Changes in amino acids can also alter post-translational processes of the antigen-binding polypeptide, such as changes in the number or location of glycosylation sites.

[0180] For example, one, two, three, four, five, or six amino acids may be inserted, substituted, or deleted in each of the CDRs (depending, of course, on their lengths), while one, two, three, four, five, six, seven, eight, nine, ten, one, two, three, three, four, nine, ten, one

[0181] The most important sites for substitutional mutagenesis are the CDRs of the heavy chain and / or light chain, particularly the hypervariable region (HDR), but modifications of the FRs in the heavy chain and / or light chain are also intended. The 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 CDR, 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 region (FR). For example, if the CDR sequence contains 6 amino acids, 1, 2, or 3 of these amino acids are expected to be substituted. Similarly, if the CDR sequence contains 15 amino acids, 1, 2, 3, 4, 5, or 6 of these amino acids are expected to be substituted.

[0182] A useful method for identifying specific residues or regions of antigen-binding polypeptides that are favorable for mutagenesis is called the "alanine scanning mutagenesis method," as described by Cunningham and Wells in Science, 244:1081-1085 (1989). In this method, residues or target residue groups within the antigen-binding polypeptide are identified (e.g., charged residues such as arg, asp, his, lys, and glu) and replaced with neutral or negatively charged amino acids (most preferably alanine or polyalanine) that influence the interaction between the amino acid and the epitope.

[0183] Next, further variants or other variants are introduced at the substitution site, i.e., in place of the substitution site, to select amino acid positions that are functionally sensitive to the substitution. Thus, the site or region to which the amino acid sequence variant is introduced is predetermined, but the nature of the mutation itself does not need to be predetermined. For example, alanine scanning or random mutagenesis may be performed at the target codon or target region to analyze or optimize the performance of the mutation at a given site, and variants of the expressed antigen-binding polypeptide are screened for the optimal combination of desired activity. Techniques for introducing substitutional mutations at predetermined sites in DNA with known sequences are well known, such as M13 primer mutagenesis and PCR mutagenesis. Screening of variants is performed using antigen-binding activity assays such as target cell surface antigen or CD3 binding.

[0184] Generally, when one or more or all of the CDRs of the heavy chain and / or light chain are substituted with amino acids, the resulting “substituted” 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 the degree to which it is identical to the “substituted” sequence depends on the length of the CDR. For example, a CDR with five amino acids is preferably 80% identical to its substituted sequence because it has at least one substituted amino acid. Therefore, the CDRs of antigen-binding polypeptides may have different degrees of identity with respect to their substituted sequences; for example, CDRL1 may have 80% identity while CDRL3 may have 90% identity.

[0185] A preferred substitution (or replacement) is a conservative substitution. However, any substitution (including non-conservative substitutions or one or more of the “exemplary substitutions” listed in Table 3 below) is conceivable, as long as the antigen-binding polypeptide retains its ability to bind to the surface antigen of the target cell via the first domain and to CD3 or CD3 epsilon via the second domain, and / or its CDR is identical to the substituted sequence (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).

[0186] Conservative substitutions are shown in Table 3 under the heading "Preferred Substitutions." If such substitutions alter biological activity, they are referred to as "Exemplary Substitutions" in Table 3, or more substantial modifications are introduced in relation to amino acid classes, as further described below, and the product can be screened for desired characteristics.

[0187] [Table 3]

[0188] Substantial modification of the biological properties of the antigen-binding polypeptide of the present invention is achieved by selecting substitutions that have a significantly different effect on (a) the structure of the polypeptide backbone of the substitution region, for example, as a sheet-like or helical three-dimensional structure, (b) the molecular charge or hydrophobicity at the target site, or (c) the maintenance of the bulkiness 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.

[0189] Non-conservative substitutions involve exchanging one member of one class for another. Substituting any cysteine ​​residue that does not contribute to maintaining the proper conformation of the antigen-binding polypeptide with serine residues can generally improve the molecular oxidative stability and avoid abnormal crosslinking. Conversely, adding cysteine ​​bonds to an antigen-binding polypeptide can improve its stability (especially if the antigen-binding polypeptide is an antibody fragment such as an Fv fragment).

[0190] With respect to 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 similarity search method of Pearson and Lipman, 1988, Proc. Nat. Acad. Sci. USA 85:2444, computer execution of these algorithms (GAP, BESTFIT, FASTA, and TFASTA from Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Drive, Madison, Wis.), the Best Fit sequence program described by Develeux et al., 1984, Nucl. Acid Res. 12:387-395, preferably with default settings, or by visual inspection. Preferably, the identity percentage is calculated by FastDB based on the following parameters: mismatch penalty of 1; gap penalty of 1; gap size penalty of 0.33; and join 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.

[0191] One example of a useful algorithm is PILEUP. PILEUP generates a multisequence alignment from related sequences using pairwise alignment with a progressive method. This also allows plotting a tree showing the clustering relationships used to generate the alignment. PILEUP uses a simplified version of the progressive alignment method described by Feng & Doolittle, 1987, J.Mol.Evol.35:351-360. This method is similar to that described by 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 a weighted end gap.

[0192] 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.USA90:5873-5787. 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 their default values. The adjustable parameters are set to the following values: overlap span = 1, overlap fraction = 0.125, word threshold (T) = II. The HSP S-parameter and HSP S2-parameter are dynamic values, established by the program itself, depending on the composition of a particular sequence and the composition of the specific database from which the target sequence is searched; however, the values ​​can be adjusted to increase sensitivity.

[0193] An additional useful algorithm is gapped BLAST, reported by Altschul et al., 1993, Nucl. Acids Res. 25:3389-3402. Gapped BLAST uses BLOSUM-62 substitution scores, with the threshold parameter T set to 9, and a two-hit method resulting in gapless extension with a cost of 10+k for a gap length k, where Xu is set to 16 and Xg is set to 40 during the database search phase and 67 during the algorithm's output phase. Gapped alignment is initiated with a score corresponding to approximately 22 bits.

[0194] Generally, the amino acid homology, similarity, or identity between individual variant CDR or VH / VL sequences is at least 60% with respect to the sequences shown herein, and more typically, the homology or identity is preferably 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, the “percentage of nucleic acid sequence identity (%)” with respect to the nucleic acid sequences of 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 polypeptide. In a specific method, the BLASTN module of WU-BLAST-2 is utilized, with the overlap span and overlap fraction set to default parameters of 1 and 0.125, respectively.

[0195] Generally, the nucleic acid sequence homology, similarity, or identity between the nucleotide sequences encoding individual variant CDR or VH / VL sequences and the nucleotide sequences shown herein is at least 60%, and more typically, it is preferable that the homology or identity increases to 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 nearly 100%. Therefore, the "Variant CDR" or "Variant VH / VL region" has specific homology, similarity or identity with respect to the parent CDR / VH / VL of the present invention and shares biological functions 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.

[0196] In one embodiment, the percentage of identity of the antigen-binding polypeptide according to the present invention to human germline is ≥70% or ≥75%, more preferably ≥80% or ≥85%, even more preferably ≥90%, most preferably ≥91%, ≥92%, ≥93%, ≥94%, ≥95%, or even further ≥96%. Identity to human antibody germline gene products is considered an important feature for reducing the risk that therapeutic proteins will induce an immune response to a drug in a patient undergoing treatment. Hwang & Foote ("Immunogenicity of engineered antibodies"; Methods 36(2005)3-10) have demonstrated that reducing the non-human portion of a drug-antibody construct leads to a reduction in the risk of inducing anti-drug antibodies in patients undergoing treatment. By comparing a vast number of clinically evaluated antibody drugs and corresponding immunogenicity data, humanization of the V region of antibodies tends to result in lower protein immunogenicity (average 5.1% of patients) compared to antibodies possessing the unmodified, non-human V region (average 23.59% of patients). Therefore, high identity to the human sequence is desirable for protein therapeutics in the form of antibody constructs based on the V region. To determine this germline identity, the V region of 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 as a percentage by dividing the number of identical amino acid residues by the total number of amino acid residues in VL. A similar method is possible for the VH segment (http: / / vbase.mrc-cpe.cam.ac.uk / ), except that VH CDR3 may be excluded due to its high diversity and lack of existing human germline VH CDR3 alignment partners. Next, recombinant technology can be used to increase sequence identity for human antibody germline genes.

[0197] In further embodiments, the bispecific antigen-binding polypeptides of the present invention exhibit high monomer yields under standard research-scale conditions, for example, in a standard two-step purification process. Preferably, the monomer yield of the antigen-binding polypeptide according to the present invention is ≥0.25 mg / L supernatant, more preferably ≥0.5 mg / L, even more preferably ≥1 mg / L, and most preferably ≥3 mg / L supernatant.

[0198] Similarly, the yield of dimeric antigen-binding polypeptide isoforms and, therefore, the monomer ratio of the antigen-binding polypeptide (i.e., monomer:(monomer+dimer)) can be determined. The productivity of monomer and dimeric antigen-binding polypeptides, as well as the calculated monomer ratios, can be obtained, for example, in an SEC purification step of the culture supernatant derived from standardized study-scale production in roller bottles. In one embodiment, the monomer ratio of the antigen-binding polypeptide is ≥80%, more preferably ≥85%, even more preferably ≥90%, and most preferably ≥95%.

[0199] In one embodiment, the antigen-binding polypeptide preferably has 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 polypeptide is determined by incubating the antigen-binding polypeptide in human plasma at 37°C for 24 hours, followed by 51This can be tested by determining the EC50 in a chromium-releasing cytotoxicity assay. Effector cells in the cytotoxicity assay may be stimulated, enriched human CD8-positive T cells. Target cells may be, for example, CHO cells transfected with the surface antigen of human target cells. An effector cell to target cell (E:T) ratio of 10:1 can be selected. The human plasma pool used for this purpose is derived from the blood of healthy donors collected using EDTA-coated syringes. The cellular components are removed by centrifugation, and the upper plasma phase is collected and then pooled. As a control, antigen-binding polypeptides are diluted in RPMI-1640 medium immediately before the cytotoxicity assay. Plasma stability is calculated as the ratio of EC50 (after plasma incubation) to EC50 (control).

[0200] A low conversion rate from monomer to dimer of the antigen-binding polypeptide of the present invention is more preferable. The conversion rate can be measured under different conditions and analyzed by high-speed size exclusion chromatography. For example, incubation of the monomer isoform of the antigen-binding polypeptide can be carried out in an incubator at a concentration of, for example, 100 μg / ml or 250 μg / ml and 37°C for 7 days. Under these conditions, the antigen-binding polypeptide of the present invention preferably exhibits a dimer ratio of ≤5%, more preferably ≤4%, even more preferably ≤3%, even more preferably ≤2.5%, even more preferably ≤2%, even more preferably ≤1.5%, most preferably ≤1%, or ≤0.5%, or even 0%.

[0201] The bispecific antigen-binding polypeptide of the present invention also preferably exhibits a very low dimerization rate after several freeze / thaw cycles. For example, the monomer of the antigen-binding polypeptide is adjusted to a concentration of 250 μg / ml in a general-purpose pharmaceutical buffer, subjected to three freeze / thaw cycles (freezing at -80°C for 30 minutes, then thawing at room temperature for 30 minutes), and then subjected to a high-speed SEC to determine the proportion of the initial monomer antigen-binding polypeptide converted to a dimer antigen-binding polypeptide. Preferably, the proportion of dimers of the bispecific antigen-binding polypeptide is, for example, ≤5%, more preferably ≤4%, even more preferably ≤3%, even more preferably ≤2.5%, even more preferably ≤2%, even more preferably ≤1.5%, most preferably ≤1%, or even further ≤0.5% after three freeze / thaw cycles.

[0202] The bispecific antigen-binding polypeptide of the present invention preferably exhibits 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. The thermal stability parameter can be determined from the viewpoint of antibody aggregation temperature as follows: A 250 μg / ml antibody solution is transferred to 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 continuously acquiring the measurement radius. The antibody aggregation temperature is calculated using the increase in radius indicating protein melting and aggregation.

[0203] Alternatively, the melting temperature curve can be measured by differential scanning calorimetry (DSC) to determine the intrinsic biophysical protein stability of the antigen-binding polypeptide. These experiments are performed using a MicroCal LLC (Northampton, MA, USA) VP-DSC instrument. Energy uptake of a sample containing the antigen-binding polypeptide is recorded from 20°C to 90°C and compared with a sample containing only the formulation buffer. The antigen-binding polypeptide is adjusted to a final concentration of, for example, 250 μg / ml in SEC running buffer. The overall temperature of the sample is increased stepwise to record each melting curve. Energy uptake of the sample and the formulation buffer standard is recorded at each temperature T. The difference in energy uptake Cp (kcal / mole / °C) obtained by subtracting the standard from the sample is plotted against each temperature. The melting temperature is defined as the temperature at which energy uptake first reaches its maximum.

[0204] The target cell surface antigen × CD3 bispecific antigen-binding polypeptide of the present invention is also assumed to have a turbidity of ≤0.2, preferably ≤0.15, more preferably ≤0.12, even more preferably ≤0.1, and most preferably ≤0.08 (measured by OD340 after concentrating the purified monomer antigen-binding polypeptide to 2.5 mg / ml and incubating overnight).

[0205] In further embodiments, the antigen-binding polypeptide according to the present invention is stable at physiological pH or slightly lower, i.e., at approximately pH 7.4 to 6.0. The higher the tolerance exhibited by the antigen-binding polypeptide at non-physiological pH, for example, approximately pH 6.0, the higher the recovery rate of the antigen-binding polypeptide eluted from the ion-exchange column relative to the total amount of loaded protein. The recovery rate of the antigen-binding polypeptide from the ion (e.g., cation) exchange column at approximately 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%.

[0206] The bispecific antigen-binding polypeptides of the present invention are further expected to exhibit therapeutic efficacy or antitumor activity. This can be evaluated, for example, in the tests disclosed in the following examples of advanced-stage human tumor xenograft models.

[0207] Those skilled in the art know how to obtain meaningful and reproducible results while modifying or adapting specific parameters of this test, such as the number of tumor cells injected, the injection site, the number of human T cells transplanted, the amount of bispecific antibody construct administered, and the schedule. Preferably, the tumor growth suppression T / C [%] is ≤70 or ≤60, more preferably ≤50 or ≤40, even more preferably ≤30 or ≤20, most preferably ≤10 or ≤5 or even further ≤2.5.

[0208] In preferred embodiments of the antigen-binding polypeptide of the present invention, the antigen-binding polypeptide is a single-chain antigen-binding polypeptide.

[0209] Furthermore, in a preferred embodiment of the antibody construct of the present invention, the third domain is arranged in the order of amino to carboxyl, Hinge-CH2-CH3-Linker-Hinge-CH2-CH3 Includes.

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

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

[0212] For example, a preferred feature of the antigen-binding polypeptide of the present invention compared to bispecific hetero-Fc antibody constructs known in the art may be related, in particular, to the introduction of the above-mentioned modification in the CH2 domain. Therefore, with respect to the polypeptide of the present invention, it is preferable that the CH2 domain in the third domain of the antigen-binding polypeptide of the present invention contains intradomain cysteine ​​disulfide crosslinks at Kabat positions 309 and 321, and / or that the glycosylation site at Kabat position 314 is removed by N314X substitution, preferably N314G substitution, as described above.

[0213] In a more preferred embodiment of the present invention, the CH2 domain in the third domain of the antigen-binding polypeptide of the present invention includes intradomain cysteine ​​disulfide crosslinks at Kabat positions 309 and 321, and the glycosylation site at Kabat position 314 is removed by N314G substitution.

[0214] In one embodiment, the present invention is an antigen-binding polypeptide, (i) Does the first domain contain two antibody-variable domains, and the second domain contain two antibody-variable domains? (ii) Does the first domain contain one antibody variable domain and the second domain contain two antibody variable domains? (iii) The first domain contains two antibody-variable domains and the second domain contains one antibody-variable domain; or (iv) The first domain contains one antibody variable domain, and the second domain contains one antibody variable domain. Provides antigen-binding polypeptides.

[0215] Therefore, the first and second domains may each be binding domains containing two antibody-variable domains, such as VH and VL domains. Examples of such binding domains containing two antibody-variable domains as described herein include, for example, the Fv fragment, scFv fragment, or Fab fragment as described herein. Alternatively, one or both of these binding domains may contain only a single variable domain. Examples of such single-domain binding domains as described herein include, for example, nanobody or single-variable-domain antibodies containing 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.

[0216] In preferred embodiments of the antigen-binding polypeptide of the present invention, the first and second domains are fused to a third domain via a peptide linker. Preferred peptide linkers are described above herein and are characterized by the amino acid sequence Gly-Gly-Gly-Gly-Ser, i.e., Gly4Ser (SEQ ID NO: 187) or a polymer thereof, i.e., (Gly4Ser)x, where x is an integer of 1 or more (e.g., 2 or 3). A particularly preferred linker for fusion of the first and second domains to the third domain is shown in SEQ ID NO: 1.

[0217] In preferred embodiments, the antigen-binding polypeptide of the present invention is composed of amino and carboxyl molecules in that order. (a) The first domain; (b) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs. 187-189; (c) Second domain; (d) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 187, 188, 189, 195, 196, 197, and 198; (e) The 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: 191, 192, 193, and 194; and (g) Second polypeptide monomer of the third domain It is characterized by including.

[0218] In one aspect of the present invention, the surface antigen of the target cell to which the first domain binds is a tumor antigen, an antigen specific to an immunodeficiency, or a viral antigen. As used herein, the term “tumor antigen” may be understood as those antigens presented on tumor cells. These antigens may be presented on the cell surface in the extracellular portion and often comprise both the transmembrane and cytoplasmic portions of the molecule. These antigens may, in some cases, be presented only by tumor cells and never by normal cells. Tumor antigens may be expressed exclusively on tumor cells or may exhibit tumor-specific mutations compared to normal cells. In this case, they are called tumor-specific antigens. More general antigens are those presented by both tumor cells and normal cells and are called tumor-associated antigens. These tumor-associated antigens may be overexpressed compared to normal cells or, due to the less compact structure of tumor tissue compared to normal tissue, are accessible to tumor cells for antibody binding. Non-exclusive examples of tumor antigens used herein include CDH19, MSLN, DLL3, FLT3, EGFRvIII, CD33, CD19, CD20, CD70, BCMA, and PSMA.

[0219] Further target cell surface antigens specific to immunodeficiency in connection with the present invention include, for example, TL1A and TNF-alpha. The targets are preferably addressed by the bispecific antigen-binding polypeptide of the present invention, which is preferably a full-length antibody. In a very preferred embodiment, the antibody of the present invention is a hetero-IgG antibody.

[0220] In preferred embodiments of the antigen-binding polypeptide of the present invention, the tumor antigen is selected from the group consisting of CDH19, MSLN, DLL3, FLT3, EGFRvIII, CD33, CD19, CD20, CD70, BCMA, and PSMA.

[0221] In one aspect of the present invention, the antigen-binding polypeptide is composed of amino and carboxyl molecules in that order. (a) Sequence numbers 7, 8, 17, 27, 28, 37, 38, 39, 40, 41, 48, 49, 50, 51, 52, 59, 60, 61, 62, 63, 64, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 89, 90, 91, 92, 93, 100, 101, 102, 103, 104, 113, 114, 121, 122, 123, 124, 12 A first domain having an amino acid sequence selected from the group consisting of 5, 131, 132, 133, 134, 135, 136, 143, 144, 145, 146, 147, 148, 149, 150, 151, 158, 159, 160, 161, 162, 163, 164, 165, 166, 173, 174, 175, 176, 177, 178, 179, 180, 181; (b) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs. 187-189; (c) A second domain having 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, or SEQ ID NO. 202 from International Publication No. 2008 / 119567; (d) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 187, 188, 189, 195, 196, 197, and 198; (e) A first polypeptide monomer of a third domain having a polypeptide sequence selected from the group consisting of Sequence IDs 17-24 of International Publication No. 2017 / 134140; (f) A peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 191, 192, 193, and 194; and (g) comprising a second polypeptide monomer of a third domain having a polypeptide sequence selected from the group consisting of Sequence IDs 17-24 of International Publication No. 2017 / 134140.

[0222] In one embodiment, the bispecific antigen-binding polypeptide of the present invention comprises the following, and is characterized by having an amino acid sequence selected from the group that is induced by the surface antigen of each target cell: (a) Sequence IDs 27, 28, 37-41; CD33 (b) Each of sequence numbers 48-52; EGFRvIII (c) Each of sequence numbers 59-64; MSLN (d) Each of sequence numbers 71-82; CDH19 (e) Each of sequence numbers 100 to 104; DLL3 (f) Sequence IDs 7, 8, 17, 113 and 114; CD19 (g) Each of sequence numbers 89-93; FLT3 (h) Each of sequence numbers 121-125; CDH3 (i) Each of sequence numbers 132-136; BCMA, and (j) Sequence numbers 143-151, 158-166, and 173-181, respectively; PSMA.

[0223] The present invention further provides polynucleotide / nucleic acid molecules encoding the antigen-binding polypeptide of the present invention. A polynucleotide is a biomacromolecule composed of 13 or more nucleotide monomers covalently linked in a chain. DNA (such as cDNA) and RNA (such as mRNA) are examples of polynucleotides having different biological functions. A nucleotide is an organic molecule that functions as a monomer or subunit of a nucleic acid molecule such as DNA or RNA. Nucleic acid molecules or polynucleotides can be double-stranded and single-stranded, linear and cyclic. It is preferably contained within a vector contained in a host cell. The host cell can then express the antigen-binding polypeptide after being transformed or transfected with, for example, the vector or polynucleotide of the present invention. For this purpose, the polynucleotide or nucleic acid molecule is operably linked to a control sequence.

[0224] The genetic code is a set of rules that translate the information encoded within genetic material (nucleic acids) into proteins. Biological decoding in living cells is carried out by ribosomes, which use tRNA molecules—which carry amino acids and read three nucleotides from mRNA at once—to link the amino acids in the order specified by the mRNA. This code defines how a sequence of three nucleotides, called a codon, specifies the next amino acid to be added during protein synthesis. With some exceptions, a three-nucleotide codon in a nucleic acid sequence specifies one amino acid. Since most genes are encoded by the exact same code, this particular code is often referred to as the reference genetic code or standard genetic code. While the genetic code determines the protein sequence of a given coding region, other genomic regions can influence when and where these proteins are produced.

[0225] Furthermore, the present invention provides vectors comprising the polynucleotide / nucleic acid molecule of the present invention. A vector is a nucleic acid molecule used as a medium for transferring (foreign) genetic material into cells. The term “vector” includes, but is not limited to, plasmids, viruses, cosmids, and artificial chromosomes. Generally, genetically engineered vectors include an origin of replication, a multicloning site, and a selection marker. The vector itself is generally a nucleotide sequence, generally a DNA sequence, containing an insert (transgene) and a larger sequence that serves as the “backbone” of the vector. Modern vectors may include, in addition to the transgene insert and backbone, the following additional features: promoters, genetic markers, antibiotic resistance, reporter genes, targeting sequences, and protein purification tags. Vectors called expression vectors (expression constructs) are specifically for the expression of a transgene in target cells and generally have regulatory sequences.

[0226] The term "regulatory sequence" refers to a DNA sequence necessary for the expression of an operablely linked coding sequence in a particular host organism. Suitable regulatory sequences for prokaryotes include, for example, promoters, optionally operator sequences, and ribosome-binding sites. Eukaryotic cells are known to utilize promoters, polyadenylation signals, and enhancers.

[0227] Nucleic acids are "operably linked" if they have a functional relationship with another nucleic acid sequence. For example, DNA for a pre-sequence or secretion leader is operably linked to DNA for a polypeptide if it is expressed as a protein precursor involved in polypeptide secretion; 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 to facilitate translation. Generally, "operably linked" means that the linked DNA sequences are contiguous, and in the case of a secretion leader, contiguous and within the read frame. Enhancers, however, do not need to be contiguous. Linking is done by ligation at a convenient restriction site. If such a site does not exist, synthetic oligonucleotide adapters or linkers are used, according to conventional practice.

[0228] "Transfection" is the process of intentionally introducing nucleic acid molecules or polynucleotides (including vectors) into target cells. This term is primarily used for non-viral methods in eukaryotic cells. Transduction is often used to describe the viral transfer of nucleic acid molecules or polynucleotides. Transfection of animal cells generally involves creating transient pores or "holes" in the cell membrane to allow material uptake. Transfection can be performed using calcium phosphate, by electroporation, by compressing the cell, or by mixing cationic lipids with liposome-generating substances, fusing them with the cell membrane, and accumulating internal cargo.

[0229] The term "transformation" is used to describe the nonviral transfer of nucleic acid molecules or polynucleotides (including vectors) into bacteria and non-animal eukaryotic cells, including plant cells. Therefore, transformation is a genetic modification of a bacterial or non-animal eukaryotic cell resulting from direct uptake from its periphery across the cell membrane and subsequent integration of exogenous genetic material (nucleic acid molecules). Transformation can be induced by artificial means. For transformation to occur, the cell or bacterium must be in a competent state where transformation can occur as a timed response to environmental conditions such as starvation and cell density.

[0230] Furthermore, the present invention provides host cells transformed or transfected with the polynucleotide / nucleic acid molecules or vectors of the present invention. As used herein, the terms “host cell” or “recipient cell” are intended to include any individual cell or cell culture that may or may have been a recipient of the vectors, exogenous nucleic acid molecules and polynucleotides encoding the antigen-binding polypeptides of the present invention; and / or the antigen-binding polypeptide itself. The introduction of each substance into a cell is carried out by transformation, transfection, etc. The term “host cell” is also intended to include single-cell offspring or potential offspring. In subsequent generations, certain modifications may occur due to spontaneous, accidental, or intentional mutations, or due to environmental influences, so such offspring may not actually be completely identical to the parent cell (morphologically or with respect to the genome or total DNA set), but this is still within the scope of the terms as used herein. Suitable host cells include, but are not limited to, prokaryotic or eukaryotic cells, as well as 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.

[0231] The antigen-binding polypeptide of the present invention can be produced in bacteria. After expression, the antibody construct of the present invention can be isolated from the E. coli cell paste in the 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 for antibodies expressed in CHO cells.

[0232] In addition to prokaryotes, eukaryotic microorganisms such as filamentous fungi or yeasts are suitable cloning or expression hosts for the antigen-binding polypeptide of the present invention. Saccharomyces cerevisiae or common baker's yeast are the most commonly used lower eukaryotic host microorganisms. However, many other genera, species, and strains are generally available and useful in the present invention, for example, Schizosaccharomyces pombe, K. lactis, K. fragilis (ATCC 12424), K. bulgaricus (ATCC 16045), K. wickeramii (ATCC 24178), K. waltii (ATCC 56500), K. drosophilarum (ATCC 16045) Hosts of the genus Kluyveromyces, such as K. thermotolerans and K. marxianus; yarrowia (European Patent No. 402226); Pichia pastoris (European Patent No. 183070); Candida; Trichoderma reesia (European Patent No. 244234); Neurospora crassa; Schwanniomyces occidentalis Hosts include the genus Schwanniomyces (such as Schwanniomyces occidentalis), as well as filamentous fungi such as Neurospora, Penicillium, Tolypocladium, and Aspergillus, such as A. nidulans and A. niger.

[0233] Suitable host cells for the expression of the glycosylated antigen-binding polypeptide of the present invention are derived from multicellular organisms. Examples of invertebrate cells include plant cells and insect cells. Many baculovirus strains and variants, as well as corresponding acceptable insect host cells derived from host organisms, have been identified (e.g., Spodoptera frugiperda (caterpillar), Aedes aegypti (mosquito), Aedes albopictus (mosquito), Drosophila melanogaster (fruit fly), and Bombyx mori (silkworm)). Various virus strains for transfection (e.g., the L-1 variant of Autographa californica NPV and the Bm-5 strain of Bombyx mori NPV) are publicly available, and such viruses may be used as the viruses of this specification according to the present invention, and in particular may be used for transfection of Spodoptera frugiperda cells.

[0234] Plant cell cultures of cotton, maize, potato, soybean, petunia, tomato, Arabidopsis, and tobacco can also be used as hosts. Cloning and expression vectors useful for protein production 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.

[0235] However, the highest interest is in vertebrate cells, and the growth of vertebrate cells under culture (tissue culture) has become a routine procedure. Examples of useful mammalian host cell lines are the simian kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the 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)); simian kidney cells (CVI ATCC CCL 70); African green monkey kidney cells (VERO-76, ATCC CRL 1587); human cervical cancer cells (HELA, ATCC CCL 2); dog 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 N.Y Acad. Sci. (1982) 383:44-68); MRC 5 cells; FS4 cells; and the human hepatoma line (Hep G2).

[0236] In a further embodiment, the invention provides a method for the production of an antigen-binding polypeptide of the invention, the method comprising culturing a host cell of the invention under conditions that permit the expression of the antigen-binding polypeptide of the invention, and recovering the produced antigen-binding polypeptide from the culture.

[0237] 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 culture medium. The term “expression” includes all steps involved in the preparation of the antigen-binding polypeptides of the present invention, such as, but not limited to, transcription, post-transcriptional modification, translation, post-translational modification, and secretion.

[0238] When recombinant technology is used, antigen-binding polypeptides can be produced in the perimembranous space within cells or secreted directly into the culture medium. If the antigen-binding protein is produced intracellularly, the first step is to remove granular fragments of host cells or lysed fragments, for example, by centrifugation or ultrafiltration. Carter et al., Bio / Technology 10:163-167 (1992) describe a procedure for isolating antibodies secreted into the perimembranous space of Escherichia coli (E. coli). Briefly, the cell paste is thawed for about 30 minutes in the presence of sodium acetate (pH 3.5), EDTA, and phenylmethylsulfonyl fluoride (PMSF). Cell fragments can be removed by centrifugation. If the antibody is secreted into the culture medium, the supernatant from such an expression system is generally concentrated first using a commercially available protein concentration filter, such as an Amicon or Millipore Pellicon ultrafiltration unit. To inhibit protein degradation, a protease inhibitor such as PMSF may be included in one of the aforementioned steps, and an antibiotic may be included to prevent the growth of exogenous contaminating bacteria.

[0239] The antigen-binding polypeptide of the present invention, prepared from host cells, can be recovered or purified using, for example, hydroxyapatite chromatography, gel electrophoresis, dialysis, and affinity chromatography. Depending on the recovered antibody, other protein purification techniques may be used, such as fractionation by ion exchange column, ethanol precipitation, reverse-phase HPLC, silica chromatography, and heparin SEPHAROSE. TMChromatography according to [reference], chromatography using an anion or cation exchange resin (e.g., a polyaspartic acid column), 59yophili-focusing, SDS-PAGE, and ammonium sulfate precipitation can also be used. When the antigen-binding polypeptide of the present invention contains a CH3 domain, Bakerbond ABX resin (J.T. Baker, Phillipsburg, NJ) is useful for purification.

[0240] Affinity chromatography is a preferred purification technique. The matrix to which the affinity ligand binds is most often agarose, but other matrices are also available. Mechanically stable matrices, such as controlled pore glass or poly(styrene divinyl)benzene, allow for faster flow rates and shorter processing times than those achievable using agarose.

[0241] Furthermore, the present invention provides a pharmaceutical composition comprising the antigen-binding polypeptide of the present invention or an antigen-binding polypeptide produced according to the method of the present invention. In the pharmaceutical composition of the present invention, the homogeneity of the antigen-binding polypeptide is preferably ≧80%, more preferably ≧81%, ≧82%, ≧83%, ≧84% or ≧85%, still more preferably ≧86%, ≧87%, ≧88%, ≧89% or ≧90%, yet still more preferably ≧91%, ≧92%, ≧93%, ≧94% or ≧95%, most preferably ≧96%, ≧97%, ≧98% or ≧99%.

[0242] As used herein, the term “pharmaceutical composition” refers 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 polypeptides of the present invention, preferably in a therapeutically effective amount. Preferably, the pharmaceutical composition further comprises one or more suitable formulations of pharmaceutically effective carriers, stabilizers, excipients, diluents, solubilizers, surfactants, emulsifiers, preservatives and / or adjuvants. The components of the acceptable composition are preferably nontoxic to the recipient at the doses and concentrations employed. Pharmaceutical compositions of the present invention include, but are not limited to, liquid, freeze-dried, and lyophilized compositions.

[0243] The compositions of the present invention may contain pharmaceutically acceptable carriers. Generally, as used herein, “pharmaceutically acceptable carriers” means any aqueous and non-aqueous solutions, sterile solutions, solvents, buffers, such as phosphate-buffered saline (PBS) solutions, water, suspensions, emulsions such as oil / water emulsions, various types of wetting agents, liposomes, dispersion media, and coatings suitable for pharmaceutical administration, particularly parenteral administration. The use of such media and agents in pharmaceutical compositions is well known in the art, and compositions containing such carriers can be formulated by well known conventional methods.

[0244] Certain embodiments provide pharmaceutical compositions comprising the antigen-binding polypeptide of the present invention and one or more excipients, such as those described exemplary in this section and elsewhere in this specification. Excipients can be used in consideration of a wide range of purposes, such as methods of the present invention for adjusting the physical, chemical or biological properties of a formulation, such as viscosity, and / or improving efficacy and / or stabilizing such a formulation, as well as methods for preventing degradation and damage caused by stress during, for example, manufacturing, transport, storage, preparation before use, administration and thereafter.

[0245] In certain embodiments, the pharmaceutical composition may contain formulation materials intended to modify, maintain, or protect, for example, the composition's pH, molar osmotic pressure, viscosity, clarity, color, isotonicity, odor, sterility, stability, dissolution or release rate, adsorption or osmosis (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: • Charged amino acids, preferably lysine, lysine acetate, arginine, glutamate and / or histidine, for example, glycine, alanine, glutamine, asparagine, threonine, proline, 2-phenylalanine; • Antimicrobial agents such as antibacterial agents and antifungal agents; • Antioxidants such as ascorbic acid, methionine, or sodium bisulfite; Buffers, buffer systems, and buffering agents used to maintain compositions at or slightly below physiological pH; examples of buffers include borates, bicarbonates, tris-HCl, citrates, phosphates or other organic acids, succinates, phosphates, and histidine; for example, Tris buffer with a pH of approximately 7.0-8.5; Non-aqueous solvents, such as propylene glycol, polyethylene glycol, vegetable oils such as olive oil, and injectable organic esters such as ethyl oleate; • Aqueous carriers containing water, alcohol / aqueous solution, emulsion, or suspension, including physiological saline and a buffer medium; • Biodegradable polymers such as polyester; • Dose extenders such as mannitol or glycine; • Chelating agents such as ethylenediaminetetraacetic acid (EDTA); • Isotonic agents and absorption retarders; • Caffeine, polyvinylpyrrolidone, beta-cyclodextrin, or hydroxypropyl-beta-cyclodextrin, or other complexing agents; • Filler; Monosaccharides; disaccharides; and other carbohydrates (such as glucose, mannose, or dextrin); 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 preferably immunoglobulins of human origin; • Colorants and fragrances; • Sulfur-containing reducing agents, such as glutathione, thioctic acid, sodium thioglycolate, thioglycerol, [alpha]-monothioglycerol, and sodium thiosulfate; • Diluent; ·emulsifier; • Hydrophilic polymers such as polyvinylpyrrolidone; • Salt-forming counterions such as sodium; Preservatives, such as antibacterial agents, antioxidants, chelating agents, and inert gases; examples include 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 (such as 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, cyclitol (e.g., inositol), polyethylene glycol; and polyhydric sugar alcohols; • Suspending agent; • Surfactants or wetting agents, such as Pluronic, PEG, sorbitan esters, polysorbates, such as polysorbate 20, polysorbate, Triton, tromethamine, lecithin, cholesterol, tyroxapole; the surfactant may preferably be a detergent having a molecular weight of >1.2 kD and / or a polyether having 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; • Stabilizers such as sucrose or sorbitol; • Isotonic enhancers, such as alkali metal halides, preferably sodium chloride or potassium chloride, mannitol, or sorbitol; Parenteral delivery vehicle containing sodium chloride solution, ringer's dextrose, dextrose and sodium chloride, lactated ringer's solution or fixative oil; • Intravenous delivery vehicles containing body fluids, nutritional supplements, and electrolyte supplements (such as those based on ringer's dextrose).

[0246] It will be apparent to those skilled in the art that different components of a pharmaceutical composition (e.g., those listed above) may have different effects, for example, amino acids may act as buffers, stabilizers and / or antioxidants, mannitol may act as bulking agents and / or isotonic enhancers, and sodium chloride may act as delivery vehicles and / or isotonic enhancers.

[0247] In a preferred embodiment of the present invention, the pharmaceutical composition is stable at approximately -20°C for at least 4 weeks. As is evident from the accompanying examples, the comparison of the quality of the antigen-binding polypeptide of the present invention with the quality of the corresponding state-of-the-art antibody construct can be tested using different systems. It is understood that these tests should conform to the "ICH Harmonised Tripartite Guideline: Stability Testing of Biotechnological / Biological Products Q5C and Specifications: Test procedures and Acceptance Criteria for Biotechnological / Biological Products Q6B" and thereby be selected to provide a stability indicator profile that results in reliable detection of changes in the identity, purity, and capacity 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 usually be assessed by two or more methods, and the resulting purity values ​​are method-dependent. For the purpose of stability testing, purity testing should be aligned with the method for determining degradation products.

[0248] The quality of the pharmaceutical composition containing the antigen-binding polypeptide of the present invention can be evaluated, for example, by analyzing the content of soluble aggregates (HMWS by size exclusion) in solution. Stability for at least 4 weeks at about -20°C is characterized by a content of less than 1.5% HMWS, preferably less than 1% HMWS.

[0249] The preferred method for analyzing product quality in this specification is size exclusion high performance liquid chromatography (SE-HPLC). SE-HPLC is typically performed using a size exclusion column and a UHPLC system, such as a Waters BEH200 size exclusion column (4.6×150 mm, 1.7 μm) and a Waters UHPLC system. The protein sample is directly injected and separated at a uniform concentration, for example, at a flow rate of 0.4 mL / min, using a phosphate buffer containing, for example, NaCl salt (the mobile phase was 100 mM sodium phosphate, 250 mM NaCl at pH 6.8), and the eluate was monitored by UV absorbance at 280 nm. Usually, about 6 μg of sample is loaded.

[0250] Trypsin Peptide Mapping for Chemical Modification Protein samples of bispecific antigen-binding polypeptides are digested using a filter-based method, for example, with a Millipore Microcon 30K device. The protein sample is added to the filter, centrifuged to remove the sample matrix, and then denatured in a 6M guanidine hydrochloride (GuHCl) buffer containing methionine (e.g., Thermo Fisher Scientific, Rockford, IL), reduced with 500mM dithiothreitol (DTT) (e.g., Sigma-Aldrich, St. Louis, MO) for 30 minutes at 37°C, and then alkylated by incubation with 500mM iodoacetic acid (IAA) (e.g., Sigma-Aldrich, St. Louis, MO) for 20 minutes in the dark at room temperature. Unreacted IAA is quenched by the addition of DTT. All of the above steps were performed on a filter. Subsequently, the sample is buffered by centrifugation to a digestion buffer (e.g., 50 mM Tris containing methionine, pH 7.8) to remove any remaining DTT and IAA. Trypsin digestion is carried out on a filter, for example, at 37°C for 1 hour using an enzyme-to-protein ratio of 1:20 (w / w). The digestion mixture is recovered by centrifugation and subsequently quenched by adding, for example, 8 M GuHCl in acetate buffer at pH 4.7.

[0251] Liquid chromatography-mass spectrometry (LC-MS) analysis is performed using an ultra-high-performance liquid chromatography (UPLC) system, such as the Thermo U-3000, directly coupled to a mass spectrometer, such as the Thermo Scientific Q-Exactive. Protein digests were separated by reverse phase using an Agilent Zorbax C18 RR HD column (2.1 × 150 mm, 1.8 μm) with a column temperature maintained at 50°C. Mobile phase A consisted of 0.020% (v / v) formic acid (FA) in water, and mobile phase B was 0.018% (v / v) FA in acetonitrile (I). Approximately 5 μg of digested bispecific antigen-binding polypeptide was injected into the column. Peptides were separated at a flow rate of, for example, 0.2 mL / min using a gradient (e.g., from 0.5% B to 36% B over 145 minutes). Eluted peptides were monitored by MS.

[0252] For peptide identification and modification analysis, data-dependent tandem MS (MS / MS) experiments are commonly used. A full scan is typically acquired, for example, in cation mode at 200–2000 m / z, followed by, for example, six data-dependent MS / MS scans to identify the peptide sequence. Quantification is based on mass spectrometry data from selected ion monitoring using the formula below.

number

[0253] Host cell protein (HCP) ELISA Microtiter plates are coated with rabbit anti-HCP immunoglobulin G (IgG) (Amgen, an internally produced antibody). After washing and blocking the plates, the test sample, control, and HCP calibration standard are added to the plates and incubated. Unbound proteins are washed from the plates, and the pooled rabbit anti-HCP IgG-biotin (Amgen, an internally produced antibody) is added to the plates and incubated. After further washing, streptavidin® horseradish peroxidase conjugate (HRP-conjugate) (e.g., Amersham-GE, Buckinghamshire, UK) is added to the plates and incubated. The plates are washed one last time, and the chromogenic substrate tetramethylbenzidine (TMB) (e.g., Kirkegaard and Perry Laboratories, Gaithersburg, MD) is added to the plates. Chromogenicity is stopped with 1M phosphate, and the optical density is measured with a spectrophotometer.

[0254] Other examples of evaluating the stability of the antigen-binding polypeptide of the present invention in the form of a pharmaceutical composition are provided in the appendix Examples 4-12. In those examples, embodiments of the antigen-binding polypeptide of the present invention are tested against different stress conditions in different pharmaceutical formulations, and the results are compared with other bispecific T-cell engagement antigen-binding polypeptides in extended half-life (HLE) forms known in the art. Generally, antigen-binding polypeptides provided in a particular FC form according to the present invention are assumed to be more stable against a wide range of stress conditions, such as temperature and light stress, compared to antigen-binding polypeptides provided in different HLE forms and antigen-binding polypeptides that do not have any HLE form (e.g., "canonical" antigen-binding polypeptides). 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 those skilled in the art will recognize, such improved stability against stresses that are difficult to avoid in clinical practice makes the antigen-binding polypeptide safer because it results in fewer degradation products in clinical practice. Consequently, the aforementioned improvement in stability means improved safety.

[0255] One embodiment provides an antigen-binding polypeptide of the present invention or an antigen-binding polypeptide prepared according to the method of the present invention for use in the prevention, treatment, or improvement of proliferative disorders, neoplastic diseases, viral diseases, or immune disorders.

[0256] The formulations described herein are useful as pharmaceutical compositions for treating, improving and / or preventing the pathological medical conditions described herein in patients in need thereof. The term “treatment” refers to both therapeutic treatment and preventive or protective measures. Treatment includes the application or administration of formulations to the body, isolated tissues or cells of a patient having a disease / disorder, symptoms of a disease / disorder, or predisposition to a disease / disorder, with the aim of curing, resolving, alleviating, mitigating, altering, correcting, improving, reversing, or influencing the disease, symptoms of a disease / disorder, or predisposition to a disease / disorder.

[0257] As used herein, the term “improvement” means an improvement in the disease state of a patient with a tumor, cancer, or metastatic cancer as described below herein, by administering the antigen-binding polypeptide according to the present invention to a target that requires it. Such improvement may also be considered as slowing or stopping the progression of the patient’s tumor, cancer, or metastatic cancer. As used herein, the term “prevention” means avoiding the onset or recurrence of a patient with a tumor, cancer, or metastatic cancer as described below herein, by administering the antigen-binding polypeptide according to the present invention to a target that requires it.

[0258] The term “disease” refers to a medical condition that would benefit from treatment with the antigen-binding polypeptide or pharmaceutical composition described herein. This includes chronic and acute disorders or diseases, including pathological conditions that make mammals susceptible to the disease in question.

[0259] A "neoplasm" is an abnormal growth of tissue, which, while not always, usually forms a mass. When it forms a mass, it is generally called a "tumor." Neoplasms or tumors can be benign, occult malignant (precancerous), or malignant. Malignant neoplasms are generally called cancers. They can usually invade and destroy surrounding tissues, forming metastases, that is, they spread to other parts, tissues, or organs of the body. Thus, the term "metastatic cancer" includes metastases to other tissues or organs other than those of the primary tumor. Lymphomas and leukemias are lymphoid neoplasms. For the purposes of this invention, they are also included in the terms "tumor" or "cancer."

[0260] As used herein, the term “immune disorder” refers to immune disorders such as autoimmune diseases, hypersensitivity, and immunodeficiency, in accordance with the general definition of the term.

[0261] In one embodiment, the present invention provides a method for treating or improving proliferative disorders, neoplastic diseases, viral diseases, or immunodeficiencies, comprising the step of administering an antigen-binding polypeptide of the present invention or an antigen-binding polypeptide prepared according to the method of the present invention to a subject in need thereof.

[0262] The terms “in need” or “in need of treatment” include subjects who already have the disorder and subjects for whom the disorder will be prevented. “In need” or “patient” includes human and other mammalian subjects receiving preventive or therapeutic treatment.

[0263] The antigen-binding polypeptides of the present invention are generally designed to suit specific routes and methods of administration, specific doses and frequencies of administration, and specific treatments for specific diseases, particularly in terms of bioavailability and persistence. The materials of the composition are preferably formulated at concentrations acceptable at the administration site.

[0264] In this specification, the singular forms “a,” “an,” and “the” refer to multiple objects unless otherwise explicitly indicated by the context. For example, a reference to “reagent” includes one or more such reagents, and a reference to “method” includes equivalent steps and methods known to those skilled in the art that may be modified for or substituted for the methods described herein.

[0265] Unless otherwise specified, the term “at least” preceding a set of elements should be understood to refer to all elements within that set. Those skilled in the art will recognize or be able to verify, through mere routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be incorporated herein.

[0266] Wherever the terms “and / or” are used in this specification, they include the meanings of “and,” “or,” and “all or any other combination of the elements connected by the terms.”

[0267] As used herein, the terms “about” or “approximately” mean 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.

[0268] The terms "less than" or "greater than" include specific numbers. For example, "less than 20" means less than or equal to 20. Similarly, "greater than" or "greater than" means greater than or equal to, or greater than or equal to, respectively.

[0269] Throughout this specification and the subsequent claims, unless the context requires otherwise, the word “comprise,” and variations such as “comprises” and “comprising,” will be understood to mean that they include the integer or process or group of integers or processes being described, but not to mean that they exclude any other integer or process or group of integers or processes being described.

[0270] As used herein, the term "contains" may also be replaced with the terms "contains" or "includes," or, as used herein, sometimes with the term "has." As used herein, "consists of" excludes any component, process or ingredient not specified in the components of the claim. As used herein, "essentially consists of" does not exclude materials or processes that do not substantially affect the basic and novel features of the claim.

[0271] In each example herein, the terms “contains,” “essentially consist of,” and “consist of” may be replaced with any of the other two terms.

[0272] It should be understood that the present invention is not limited to, and therefore may vary from, the specific methodologies, protocols, materials, reagents, and substances described herein. The technical terms used herein are solely for the purpose of describing specific embodiments and are not intended to limit the scope of the invention, which is defined solely by the claims.

[0273] All publications and patents (including all patents, patent applications, scientific publications, manufacturer specifications, instructions, etc.) referenced throughout this specification, whether above or below, are incorporated herein by reference in their entirety. Nothing in this specification should be construed as an acknowledgment that the present invention has no prior rights to such disclosure by prior art. This specification shall prevail over any material incorporated by reference unless such material is inconsistent with or in agreement with this specification. [Examples]

[0274] A better understanding of the present invention and its advantages can be obtained from the following examples, but these examples are provided for illustrative purposes only. These examples are not intended in any way to limit the scope of the present invention.

[0275] Example 1: Evaluation of CD33×CD3 bispecific antigen-binding polypeptide chromatographic capture using TOYOPEARL® AF-r protein L-650F compared to Capto® L. a) Column details Two pre-packed columns [lot number 65PLFC501A, part number 0045162, serial numbers 00023 and 00042] were used. The columns had an ID of 8 mm and a floor height of 10 cm. Each column had a volume of 5 mL. b) Details of the resin Two columns pre-filled with TOYOPEARL AF-r Protein L-650F resin were used. c) Pre-packed columns were used. d) Supply conditions The frozen supply solution was thawed in a 25°C water bath on the day of the test or the day before the test [held overnight at 2-8°C]. Once the supply solution reached room temperature, it was filtered and sterilized before use in the test. Results of Example 1: CD33 × CD3 bispecific antigen-binding polypeptide Two columns, each pre-filled with 5 ml of TOYOPEARL AF-r Protein L-650F resin, were linked together to achieve a total bed height of 20 cm, representing a pilot-scale bed height. Dynamic binding capacity testing was performed by flowing the load material under the conditions shown in Table 5. The achieved elution binding capacity was 12.7 g / L of the packed resin, which is a four-fold improvement over the current affinity resin. The overall yield was within the same range as the current method. Several other advantages are possible due to the four-fold improvement in binding capacity. Using the new TOYOPEARL AF-r Protein L-650F with a four-fold improvement in binding capacity would result in a six-fold reduction in the number of cycles with a reduced volume of collected cell culture medium [assuming a 5 KL supply solution, Table 5]. These are significant advantages at the production scale [Table 5].

[0276] [Table 4]

[0277] [Table 5]

[0278] Example 2: Evaluation of CD19×CD3 bispecific antigen-binding polypeptide chromatographic capture using TOYOPEARL® AF-r protein L-650F compared to Capto® L. a) Column details Only one pre-packed column with the same details as described above was used. b) Details of the resin A single column pre-filled with TOYOPEARL AF-r Protein L-650F resin was used. c) Pre-packed columns were used. d) Supply conditions The frozen supply solution was thawed in a 25°C water bath on the day of the test or the day before the test [held overnight at 2-8°C]. Once the supply solution reached room temperature, it was filtered and sterilized before use in the test. Results of Example 2: CD19 × CD3 bispecific antigen-binding polypeptide A single pre-packed column, filled with 5 ml of TOYOPEARL AF-r Protein L-650F resin and having a bed height of 10 cm representing the bed height at pilot scale, was used for binding capacity measurement. Dynamic binding capacity tests were performed by flowing the load material under the conditions shown in Table 6. The achieved elution and binding capacity was comparable to the current affinity resin, but a twofold improvement over the current affinity resin is possible. Several other advantages are possible with the twofold improvement in binding capacity. Using the new TOYOPEARL AF-r Protein L-650F with twofold improved binding capacity would result in a twofold reduction in the number of cycles with a reduced volume of collected cell culture medium [assuming a 1 KL supply solution, Table 7]. These are significant advantages at the production scale [Table 7]. Table 8 shows a comparison of product quality between a screening run and a large-scale GMP run performed using the current Capto L resin.

[0279] [Table 6]

[0280] [Table 7]

[0281] [Table 8]

[0282] Example 3: Evaluation of BCMA x CD3 bispecific antigen-binding polypeptide chromatographic capture using TOYOPEARL® AF-r protein L-650F compared to Capto® L. a) Column details A single Omnifit glass bore column with an ID of 6 mm was used, manually filled to a floor height of 5 cm. b) Details of the resin The Omnifit 6mm ID column was manually filled using a 100ml bottle of TOYOPEARL AF-r Protein L-650F resin [Lot No. 65PLFC03C]. c) Column packing In Example 3, the desired amount of TOYOPEARL AF-r Protein L-650F resin was suspended in a graduated cylinder, and the slurry percentage of the resin in the shipping buffer was calculated. The amount of resin calculated based on a specific compressibility coefficient was then transferred to a 6 mm ID Omnifit glass bore column. The resin was then fluid-filled into a 100 mM sodium chloride solution to a final target floor height of 5 cm. d) Supply conditions The frozen supply solution was thawed in a 25°C water bath on the day of the test or the day before the test [held overnight at 2-8°C]. Once the supply solution reached room temperature, it was filtered and sterilized before use in the test. Results from Example 3: BCMA × CD3 bispecific antigen-binding polypeptide In Experiment 3, the elution and binding capacity was measured under the conditions shown in Table 9 by manually loading a 6.6 mm ID Omnifit glass bore column. The achieved elution and binding capacity was comparable to that of the current affinity resin, but with a significant time saving of approximately 3 hours in the pilot-scale loading process [Table 9].

[0283] [Table 9]

[0284] [Table 10]

[0285] [Table 11]

[0286] Table 12

[0287] Table 13

[0288] Table 14

[0289] Table 15

[0290] Table 16

[0291] Table 17

[0292] Table 18

[0293] Table 19

[0294] Table 20

[0295] Table 21

[0296] Table 22

[0297] Table 23

[0298] Table 24

[0299] Table 25

[0300] Table 26

[0301] Table 27

[0302] Table 28

[0303] Table 29

[0304] Table 30

[0305] Table 31

[0306] Table 32

[0307] Table 33

[0308] Table 34

[0309] Table 35

[0310] Table 36

[0311] Table 37

[0312] Table 38

[0313] Table 39

[0314] Table 40

[0315] Table 41

Claims

1. 1. A method for purifying a bispecific antigen-binding polypeptide comprising a first domain that binds to a cell surface antigen and a second domain that binds to an extracellular epitope of the human and macaque (Macaca) CD3 epsilon chain, comprising: (a) providing a separation resin comprising a polymer matrix portion and a ligand portion, wherein the matrix portion comprises a polymer, preferably polymethacrylate, and has a particle size of at least 10 μm, preferably at least 20 μm, more preferably about 30-60 μm, and the ligand portion comprises recombinant protein L, wherein the protein L of the ligand portion is covalently bound to particles of the matrix portion; (b) contacting a process fluid containing the bispecific antigen-binding polypeptide with the separation resin; (c) capturing the bispecific antigen-binding polypeptide by the ligand moiety of the separation resin, wherein the bispecific antigen-binding polypeptide reversibly binds to the ligand moiety of the separation resin, and wherein the remainder of the process fluid does not bind to the ligand moiety of the separation resin; (d) washing the bound bispecific antigen-binding polypeptide with a wash buffer that does not elute the bispecific antigen-binding polypeptide from the ligand moiety; and (e) eluting the bispecific antigen-binding polypeptide from the ligand moiety with an elution buffer at an acidic pH A method comprising:

2. The method of claim 1 , wherein the matrix portion has a particle size of about 45 μm.

3. The method of claim 1, wherein the recombinant Protein L comprises a modified B4 domain having an alkaline-stable tetrameric ligand with multiple coupling sites.

4. The method of claim 1 , wherein the recombinant Protein L reversibly binds to a kappa light chain outside the antigen-binding site of the bispecific antigen-binding polypeptide.

5. 2. The method of claim 1, wherein the process fluid is passed through the separation resin at least once (purification cycle) to allow contact of the bispecific antigen-binding polypeptide with the Protein L (residence time), wherein the bispecific antigen-binding polypeptide residence time before elution is at least about 2 minutes, preferably about 2.5 to 4 minutes.

6. 2. The method of claim 1, wherein the wash buffer comprises at least one compound selected from the group consisting of phosphate buffered saline (PBS), preferably in a concentration range of 0.01 to 1x, 3-(N-morpholino)propanesulfonic acid (MOPS), preferably in a range of 0 to 30 mM, NaCl, preferably in a range of 50 to 150 mM, Tris, preferably in a range of 15 to 35 mM, arginine, preferably in a range of 0.25 to 1 M, and acetate, preferably in a range of 40 to 60 mM, and has a pH in the range of 5 to 8.

7. 2. The method of claim 1, wherein the elution buffer comprises at least one compound selected from the group consisting of Tris, preferably in the range of 15 to 35 mM, arginine, preferably in the range of 0.25 to 1 M, glycine, preferably in the range of 50 to 150 mM, and acetate, preferably in the range of 50 to 150 mM, and has a pH in the range of about 3 to 7.5, preferably a pH of 3.3 to 4.

2.

8. 2. The method of claim 1, wherein the dynamic loading capacity is at least 10 mg / ml resin, preferably at least 15 mg / ml resin, more preferably at least 18 mg / ml resin.

9. 2. The method of claim 1, wherein the elution binding capacity is at least 7.5 mg / ml resin, preferably at least 9 mg / ml resin, more preferably at least 16 mg / ml resin.

10. 2. The bispecific antigen-binding polypeptide of claim 1, wherein the antigen-binding polypeptide is a single-chain antigen-binding polypeptide.

11. 2. The bispecific antigen-binding polypeptide of claim 1, further comprising a third domain comprising two polypeptide monomers each comprising a hinge, a CH2 domain and a CH3 domain, wherein the two polypeptide monomers are fused to each other via a peptide linker.

12. the third domain is composed of, in order from amino to carboxyl: Hinge-CH2-CH3-linker-hinge-CH2-CH3 12. The bispecific antigen-binding polypeptide of claim 11 , comprising:

13. 12. The bispecific antigen-binding polypeptide of claim 11 , wherein each of the polypeptide monomers in 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: 203-210.

14. 12. The bispecific antigen-binding polypeptide of claim 11, wherein each of the polypeptide monomers has an amino acid sequence selected from SEQ ID NOs: 203-210.

15. 13. The bispecific antigen-binding polypeptide of claim 12, wherein the CH2 domain comprises an intradomain cysteine ​​disulfide bridge.

16. (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.

17. 2. The bispecific antigen-binding polypeptide of claim 1, wherein the first and second domains are fused to the third domain via a peptide linker.

18. The polypeptide comprises, in order from amino to carboxyl: (a) a first domain; (b) a peptide linker preferably having an amino acid sequence selected from the group consisting of SEQ ID NOs: 187-189; (c) the second domain 2. The bispecific antigen-binding polypeptide of claim 1, comprising:

19. The polypeptide comprises, in order from amino to carboxyl: (d) a peptide linker having an amino acid sequence selected from the group consisting of SEQ ID NOs: 187, 188, 189, 195, 196, 197 and 198; (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: 191, 192, 193, and 194; and (g) the second polypeptide monomer of the third domain.

18. The bispecific antigen-binding polypeptide of claim 17, further comprising:

20. 2. The bispecific antigen-binding polypeptide of claim 1, wherein the first domain of the polypeptide binds to an epitope of CD33, CD19, BCMA, PSMA, EGFRvIII, MUC17, FLT3, CD70, DLL3, CDH3 or EpCAM, preferably CD33.

21. the first binding domain comprises: (a) CDR-H1 as set forth in SEQ ID NO: 1, CDR-H2 as set forth in SEQ ID NO: 2, CDR-H3 as set forth in SEQ ID NO: 3, CDR-L1 as set forth in SEQ ID NO: 4, CDR-L2 as set forth in SEQ ID NO: 5, and CDR-L3 as set forth in SEQ ID NO: 6; (b) CDR-H1 as set forth in SEQ ID NO:29, CDR-H2 as set forth in SEQ ID NO:30, CDR-H3 as set forth in SEQ ID NO:31, CDR-L1 as set forth in SEQ ID NO:34, CDR-L2 as set forth in SEQ ID NO:35, and CDR-L3 as set forth in SEQ ID NO:36; (c) CDR-H1 as set forth in SEQ ID NO: 42, CDR-H2 as set forth in SEQ ID NO: 43, CDR-H3 as set forth in SEQ ID NO: 44, CDR-L1 as set forth in SEQ ID NO: 45, CDR-L2 as set forth in SEQ ID NO: 46, and CDR-L3 as set forth in SEQ ID NO: 47; (d) CDR-H1 as set forth in SEQ ID NO:53, CDR-H2 as set forth in SEQ ID NO:54, CDR-H3 as set forth in SEQ ID NO:55, CDR-L1 as set forth in SEQ ID NO:56, CDR-L2 as set forth in SEQ ID NO:57, and CDR-L3 as set forth in SEQ ID NO:58; (e) CDR-H1 as set forth in SEQ ID NO: 65, CDR-H2 as set forth in SEQ ID NO: 66, CDR-H3 as set forth in SEQ ID NO: 67, CDR-L1 as set forth in SEQ ID NO: 68, CDR-L2 as set forth in SEQ ID NO: 69, and CDR-L3 as set forth in SEQ ID NO: 70; (f) CDR-H1 as set forth in SEQ ID NO: 83, CDR-H2 as set forth in SEQ ID NO: 84, CDR-H3 as set forth in SEQ ID NO: 85, CDR-L1 as set forth in SEQ ID NO: 86, CDR-L2 as set forth in SEQ ID NO: 87, and CDR-L3 as set forth in SEQ ID NO: 88; (g) CDR-H1 as set forth in SEQ ID NO: 94, CDR-H2 as set forth in SEQ ID NO: 95, CDR-H3 as set forth in SEQ ID NO: 96, CDR-L1 as set forth in SEQ ID NO: 97, CDR-L2 as set forth in SEQ ID NO: 98, and CDR-L3 as set forth in SEQ ID NO: 99; (h) CDR-H1 as set forth in SEQ ID NO: 105, CDR-H2 as set forth in SEQ ID NO: 106, CDR-H3 as set forth in SEQ ID NO: 107, CDR-L1 as set forth in SEQ ID NO: 109, CDR-L2 as set forth in SEQ ID NO: 110, and CDR-L3 as set forth in SEQ ID NO: 111; (i) CDR-H1 as set forth in SEQ ID NO: 115, CDR-H2 as set forth in SEQ ID NO: 116, CDR-H3 as set forth in SEQ ID NO: 117, CDR-L1 as set forth in SEQ ID NO: 118, CDR-L2 as set forth in SEQ ID NO: 119, and CDR-L3 as set forth in SEQ ID NO: 120; (j) CDR-H1 as set forth in SEQ ID NO: 126, CDR-H2 as set forth in SEQ ID NO: 127, CDR-H3 as set forth in SEQ ID NO: 128, CDR-L1 as set forth in SEQ ID NO: 129, CDR-L2 as set forth in SEQ ID NO: 130, and CDR-L3 as set forth in SEQ ID NO: 131; (k) CDR-H1 as set forth in SEQ ID NO: 137, CDR-H2 as set forth in SEQ ID NO: 138, CDR-H3 as set forth in SEQ ID NO: 139, CDR-L1 as set forth in SEQ ID NO: 140, CDR-L2 as set forth in SEQ ID NO: 141, and CDR-L3 as set forth in SEQ ID NO: 142; (l) CDR-H1 as set forth in SEQ ID NO: 152, CDR-H2 as set forth in SEQ ID NO: 153, CDR-H3 as set forth in SEQ ID NO: 154, CDR-L1 as set forth in SEQ ID NO: 155, CDR-L2 as set forth in SEQ ID NO: 156, and CDR-L3 as set forth in SEQ ID NO: 157; and (m) CDR-H1 as set forth in SEQ ID NO: 167, CDR-H2 as set forth in SEQ ID NO: 168, CDR-H3 as set forth in SEQ ID NO: 169, CDR-L1 as set forth in SEQ ID NO: 170, CDR-L2 as set forth in SEQ ID NO: 171, and CDR-L3 as set forth in SEQ ID NO: 172; 2. The bispecific antigen-binding polypeptide of claim 1, comprising a VH region comprising CDR-H1, CDR-H2 and CDR-H3 selected from:

22. A pharmaceutical composition comprising a bispecific binding construct according to any one of claims 1 to 21.

23. 22. An antigen-binding polypeptide according to any one of claims 1 to 21 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.

24. 10. A method for improving the yield of a manufacturing process for a bispecific antigen-binding polypeptide, wherein the method of claim 1 is applied in a downstream process.

Images