Bispecific antibody effective for patients with malignant tumors other than t-cell tumors

Abstract

The present invention addresses the problem of providing a new tumor treatment method that is capable of inducing cytotoxicity to tumor cells and is unlikely to cause immunodeficiency and cytokine release syndrome. The present invention provides a therapeutic agent for tumors (tumors excluding T-cell tumors), the therapeutic agent comprising a bispecific antigen-binding molecule. The bispecific antigen-binding molecule includes (1) at least one portion that specifically binds to a target tumor antigen expressed by tumor cells, and (2) at least one portion that specifically binds to a normal T-cell-side target antigen having subtypes. The target tumor antigen expressed by the tumor cells is not present in normal T-cells. Even if the target tumor antigen is present in normal T-cells, the normal T-cells are substantially not activated when the bispecific antigen-binding molecule binds to the same antigen as the target tumor antigen present in the normal T-cells. The normal T-cells are activated by having the bispecific antigen-binding molecule bind to the normal T-cell-side target antigen. In order to provide a sufficient number of activated T-cells for the treatment of a tumor, the subtypes of the normal T-cell-side target antigen are present at sufficient proportions.

Description

The anti-TRBC2 antibody used for the TRBC2-binding portion of the bispecific antigen-binding molecule of the present invention can increase cytotoxic T cells in the presence of IL-2. Specifically, as shown in Reference Example 2 described below, the proportion of the effector memory T cells significantly increased from about 27% to about 81% on Day 15 in the presence of anti-TRBC2 antibody and IL-2. Based on this result, it was considered that stimulation with an anti-TRBC2 antibody can increase T cells that induce cytotoxicity. It is considered that stimulation with a bispecific antibody having the same antigen-binding portion as the anti-TRBC2 antibody would similarly increase T cells that induce cytotoxicity.

[0153] The anti-TRBC2 antibody used for the TRBC2-binding portion of the bispecific antigen-binding molecule of the present invention can activate T cells. Specifically, as shown in Reference Example 2 described below, stimulation with anti-TRBC2 antibody increased the ratios of CD69, Granzyme B, and Perforin-positive cells, suggesting that TRBC2-positive cells were activated by the anti-TRBC2 antibody. A bispecific antibody having the same antigen-binding portion as the anti-TRBC2 antibody is also expected to similarly have a T cell-activating effect.

[0154] Stimulation with the anti-TRBC2 antibody used for the TRBC2-binding portion of the bispecific antigen-binding molecule of the present invention did not significantly alter the ratio of PD-1-positive cells, as shown in the reference example described below. This suggested that T cells are less likely to be exhausted by activation of T cells with an anti-TRBC2 antibody. It is considered that T cells are similarly unlikely to be exhausted by a bispecific antibody having the same antigenbinding portion as the anti-TRBC2 antibody.

[0155] The anti-TRBC2 antibody used for the TRBC2-binding portion of the bispecific antigen-binding molecule of the present invention could induce T cell activation and release of cytokines, as shown in Reference Example 2 described below, but the amount of released IL-6, an inflammatory cytokine, was very small. This result indicated that the risk of cytokine release syndrome associated with anti-TRBC2 antibodies may be low. A low risk of cytokine release syndrome is similarly suggested also for a bispecific antibody that has the same antigen-binding portion as the anti- TRBC2 antibody.

[0156] The bispecific antigen-binding molecule of the present invention having a TRBC1-binding portion or TRBC2-binding portion is used in a method for (i) inducing cytotoxicity against tumor cells, (ii) increasing the ratio of TRBC1-positive cells or increasing the ratio of TRBC2-positive cells, (iii) increasing cytotoxic T cells, (iv) activating T cells, (v) stimulating or enhancing cellular responses without exhausting T cells, (vi) stimulating or enhancing cellular responses without increasing the risk of cytokine release syndrome, (vii) treating a malignancy, (viii) delaying the progression of a malignancy, or (ix) prolonging the survival of a subject suffering from a malignancy.

[0157] The bispecific antigen-binding molecule of the present invention can have antitumor activity. The antitumor activity can be measured, for example, in vivo. The antitumor activity can be evaluated, for example, as follows. For example, 1 x 107 cells / animal of human PBMCs are intravenously transplanted into immunocompromised NSG mice (Jackson Laboratory Japan), and 7 to 10 days later, 1 x 107 cells / animal of cells of tumor cell line stably expressing luciferase are intravenously transplanted. Then, the bispecific antigen-binding molecule of the present invention is administered at a dose of 0.5 mg / kg twice / week. Two to three weeks after the start of the administration, 150 mg / kg of luciferin is intraperitoneally administered to the mice, and the in vivo luciferase activity is quantified by using IVIS Imaging System (Revvity) or the like to measure the tumor volume. The anti-tumor activity can be evaluated by comparing the tumor volumes of the control group and the group administered with the bispecific antigen-binding molecule of the present invention. Examples

[0158] <Example 1> Preparation of bispecific antibody 1.1 Bispecific antibody In this example, bispecific antibodies that bind to TRBC1 or TRBC2 and a target tumor antigen were prepared. TRBC1- or TRBC2-binding portion The TRBC1-binding portion was prepared on the basis of the sequence information of the anti-TRBC1 antibody JOVI-1 clone described in U.S. Patent No. 10730942 (US10730942, Protein-based T-cell receptor knockdown, SEQ ID NOs: 15 and 16). The TRBC2-binding portion was prepared on the basis of a VH domain derived from the VH domain of hJOVI-1 described in WO2020 / 089644 (Binding Domain, SEQ ID NO: 1 in WO2020 / 089644) by three amino acid substitutions T28K, Y32F and A100N (Table 1 in WO2020 / 089644 (page 23)) and the VL domain of hJOVI-1 (SEQ ID NO: 2 in WO2020 / 089644) (the VH and VL amino acid sequences described as SEQ ID NOs: 3 and 4, respectively, in the present specification).

[0159] The portions that bind to target tumor antigens expressed on tumor cells were prepared on the basis of sequence information disclosed in the references listed in the following table. [Table 3] Target Tumor Antigen Antibody Name Reference CD19 Blinatumomab US Patent US7635472 B2 (VH: SEQ ID NO: 65, VL: SEQ ID NO: 67) BCMA Teclistamab US Patent US10072088 B2 (VH: SEQ ID NO: 27, VL: SEQ ID NO: 28) CD33 Lintuzumab US Patent US5585089 (VH: SEQ ID NO: 57, VL: SEQ ID NO: 55) CD123 Vibecotamab US Patent US9850320 B2 (VH: SEQ ID NO: 439, VL: SEQ ID NO: 462) HER2 Trastuzumab US Patent US5821337 B2 (VH: SEQ ID NO: 42, VL: SEQ ID NO: 41)

[0160] The sequences of the heavy and light chains of the bispecific antibodies prepared in this example are shown below. [Table 4] Antibody Name Tumor Antigen T cell Antigen Heavy chain Amino Acid Sequence of the Bispecific Antibody Light chain Amino Acid Sequence of the Bispecific Antibody SEQ ID NO: SEQ ID NO: Blinatumomab (scFv-scFv type) CD19 CD3 71 — CD190001 (Fab-scFv type) CD19 TRBC1 76 77 BCMA0001 (Fab-scFv type) BCMA TRBC1 78 79 CD330001 (Fab-scFv type) CD33 TRBC1 80 81 CD330003 (Fab-scFv-Fc type) CD33 TRBC1 82 83 CD1230001 (Fab-scFv type) CD123 TRBC1 84 85 CD1230003 (Fab-scFv-Fc type) CD123 TRBC1 86 87 HER20001 (Fab-scFv type) HER2 TRBC1 88 89 Schematic diagrams of the Fab-scFv, Fab-scFv-Fc, and Fab-Fc-scFv type antibodies are shown in Fig. 1.

[0161] 1.2 Preparation and purification of recombinant antibody The bispecific antibodies listed in Table 4 were constructed by expressing the heavy and light chains listed in Table 4, together with signal sequences added to them, with expression vectors, respectively. Details of the sequences of the heavy and light chains of each bispecific antibody are as described in Figs. 3A to 3G. Specifically, bispecific antibody (heavy and light chains) expression vectors were constructed by using DNA encoding each bispecific antibody and an expression vector (pcDNA3.4, Thermo Fisher Scientific). The combinations of the heavy and light chains of the bispecific antibodies are as listed in Table 4. The above expression vectors were mixed at a ratio of 1:1 and transfected into CHO cells at a DNA amount of 0.8 ^g per mL of culture medium by using ExpiFectamine CHO Reagent (Thermo Fisher Scientific). ExpiCHO Feed and ExpiFectamine CHO Enhancer were added on the next day and after 5 days of the transfection, and culture was performed at 32°C for 10 to 12 days. The cells were removed from the culture medium by centrifugation and filtration, and the culture supernatant was collected. Purification of the antibodies was performed by a combination of affinity chromatography using nickel-charged agarose and gel filtration chromatography or a combination of Protein A affinity chromatography and gel filtration chromatography.

[0162] <Example 2> Preparation of cells In this example, preparation of cells used in the experiments will be described. The following reagents were used RPMI 1640 medium (Nacalai Tesque, 30264-56) FBS (fetal bovine serum, Nichirei Bioscience, 175012, proteins were deactivated by incubation in a water bath at 56°C for 30 minutes) Penicillin-streptomycin mixed solution (Nacalai Tesque, 2625384) MEM nonessential amino acid solution 100X (Gibco, 11140-050) Sodium pyruvate solution 100 mM (Gibco, 11360-070) Anti-TRBC1 antibody (prepared and purified by the methods described in Example 1, the heavy and light chain sequences of the anti-TRBC1 antibody used are the sequences of SEQ ID NOs: 72 and 73). Imunace Injection 35 (IL-2 (interleukin 2), Shionogi Kyowa Pharmaceutical Industry, dissolved in PBS at 250 IU / pL and stored at -25°C) PBS (phosphate buffered saline, Gibco, 10010-023) Human PBMCs (peripheral blood mononuclear cells, CTL, CTL-UP1) The various cells were prepared as follows. T-LAK is a T cell population that expresses a lymphokine-activated killer (LAK) function. T-LAK was prepared with an experimental system optimized according to Ochoa AC et al. (Lymphokine-activated killer activity in long-term cultures with anti-CD3 plus interleukin 2: identification and isolation of effector subsets. Cancer Res. 1989 Feb 15;49(4):963-8). Specifically, human PBMCs were thawed in a 37°C water bath and washed once with the culture medium. After 2 days of culture in the RPMI 1640 medium containing 10% FBS supplemented with 10 nM anti-TRBC1 antibody and 100 IU / mL IL-2, the cells were further cultured in the RPMI 1640 medium containing IL-2 and 10% FBS (up to 17 days) and thereby induced into T-LAK. The cells were subcultured every 2 or 3 days, and 100 IU / mL IL-2 was newly added at the time of culture medium change.

[0163] <Reference Example 1> Verification of effector activity of anti-TRBC1 antibody and quantification of released cytokine Reagents used Flow Cytometry Staining Buffer (eBioscience, 00-4222-26) Clear Back (MBL, MT-001) Fixable Viability Dye eFluor 780 (Invitrogen, 65-0865-14) Transcription Factor Buffer Set (BD, 562574) TF Fix / Perm Buffer (4x) (51-9008100), diluted 4 times with TF Diluent Buffer before use TF Diluent Buffer (51-9008101) TF Perm / Wash Buffer (5x) (51-9008102), diluted 5 times with distilled water before use MILLIPLEX Human Cytokine / Chemokine / Growth Factor Panel A (Millipore, HCYTA-60K)

[0164] The antibodies for immunostaining are shown in the following table. [Table 5] Antibody Dye-labeled Manufacturer Item number Remarks anti-TRBC1 antibody HiLyte Fluor 647 Prepared in-house — anti-CD3 antibody FITC Invitrogen 11-0037-42 anti-CD27 antibody PE BioLegend 356406 anti-CD45RA antibody BV421 BioLegend 304130 anti-CD69 antibody PE BioLegend 310906 anti-CD25 antibody BV421 BioLegend 302630 anti-Granzyme B antibody PE BD 561142 anti-Perforin antibody BV421 BioLegend 308122 anti-PD-1 antibody BV711 BioLegend 329928 human IgG antibody HiLyte Fluor 647 Jackson Immuno Research 009-000-003 Isotype control mouse IgG2a antibody FITC BioLegend 400209 Isotype control mouse IgG1 antibody PE BioLegend 400114 Isotype control mouse IgG1 antibody BV421 BioLegend 400158 Isotype control mouse IgG2a antibody BV421 BioLegend 400259 Isotype control mouse IgG2b antibody BV421 BioLegend 400342 Isotype control mouse IgG1 antibody BV711 BioLegend 400168 Isotype control *BV421, BV711: Abbreviations for Brilliant Violet 421 (registered trademark) and Brilliant Violet 711 (registered trademark)

[0165] The anti-TRBC1 antibody was prepared and purified by the methods described in Example 1 and then dye-labeled by using HiLyte Fluor 647 Labeling Kit-NH2 (Dojindo, LK15). The heavy and light chain sequences of the anti-TRBC1 antibody are the sequences of SEQ ID NOs: 72 and 73. Human IgG (ChromPure Human IgG, whole molecule, Jackson Immuno Research, 009-000-003) was similarly dye-labeled as an isotype control. Methods The experiment was performed with reference to WO2021 / 173896 A1. Method for cell surface staining Human PBMCs or T-LAK cells were centrifuged (300 g x 5 minutes), then the supernatant was removed, and Clear Back was added. After incubation at room temperature for 5 minutes, various antibodies were added in Flow Cytometry Staining Buffer, and allowed to react at 4°C for 30 minutes. After the reaction, the cells were washed twice by adding Flow Cytometry Staining Buffer, centrifuging (500 g x 2 minutes) the suspension, and discarding the supernatant, and flow cytometric analysis was performed by using BD LSRFortessa X-20 (BD Biosciences) to quantify signals. Flow Jo_v10.6.1 was used as the analysis software.

[0166] Method for intracellular staining Human PBMCs or the T-LAK cells were centrifuged (300 g x 5 minutes), then the supernatant was removed, and Clear Back was added. After incubation at room temperature for 5 minutes, anti-CD3 and anti-TRBC1 antibodies were added in Flow Cytometry Staining Buffer, and allowed to react at 4°C for 30 minutes. After the reaction, the cells were washed twice by adding Flow Cytometry Staining Buffer, centrifuging (500 g x 2 minutes) the cell suspension, and discarding the supernatant. TF Fix / Perm Buffer contained in the Transcription Factor Buffer Set was added to the cells, and incubation was performed on ice for 30 minutes to fix and permeabilize the cells. After centrifugation (700 g x 2 minutes), the cells were washed twice by adding TF Perm / Wash Buffer, centrifuging (700g x 2 minutes) the cell suspension, and discarding the supernatant. Anti-Granzyme B or anti-Perforin antibody was added in TF Perm / Wash Buffer, and the reaction was allowed for 30 minutes on ice. After the reaction, the cells were washed twice by adding TF Perm / Wash Buffer, centrifuging (700 g x 2 minutes) the cell suspension, and discarding the supernatant, and flow cytometric analysis was performed by using BD LSRFortessa X-20 (BD Biosciences) to quantify signals. Flow Jo_v10.6.1 was used as the analysis software.

[0167] Method for quantifying released cytokines After culturing the T-LAK cells for 24 hours with the anti-TRBC1 antibody, the amounts of various cytokines released were quantified by using MILLIPLEX Human Cytokine / Chemokine / Growth Factor Panel A. The T-LAK cells were washed once with the RPMI 1640 medium (culture medium) containing 10% FBS, and then the appropriate number of the cells was counted and added to a 384-well plate. The anti-TRBC1 antibody adjusted to a final concentration of 0, 0.1, 1.0, or 10 nM in the culture medium was added to the wells in which the T-LAK cells were prepared. The 384-well plate was incubated with 5% CO2, at 37°C for 24 hours, and centrifuged (300 g, 5 minutes), and then the supernatant was collected. IFN-y, IL-2, IL-6, IL-10, and TNF-a were quantified by using MILLIPLEX Human Cytokine / Chemokine / Growth Factor Panel A. Quantified values were automatically calculated on the Luminex (registered trademark) 200 xPONET (registered trademark) 3.1 system.

[0168] Results TRBC1-positive cell ratio Figs. 4A and 4B show the increase in the ratio of TRBC1-positive cells after the addition of the anti-TRBC1 antibody. When 10 nM anti-TRBC1 antibody was added to human PBMCs and the cells were cultured in the presence of IL-2, the ratio of TRBC1-positive cells increased from 30% on Day 0 to over 80% on Day 15. It was suggested that the addition of anti-TRBC1 antibody selectively proliferated TRBC1- positive cells. It is considered that addition of a bispecific antibody having the same antigen-binding portion as the anti-TRBC1 antibody used in this experiment would similarly selectively proliferate TRBC1-positive cells.

[0169] Increase in cytotoxic T cells As shown in Figs. 5A and 5B, the T-LAK cells were stained with CD27 and CD45RA on Day 0 and Day 15, and the cell ratios of the effector memory, central memory, naive, and EMRA fractions were calculated. As a result, on Day 15, the ratio of the effector memory cells significantly increased to about 70% from about 20%, while the ratios of the other fractions decreased. On the basis of these results, it was considered that the number of T cells that induce cytotoxicity can be increased by stimulation with anti-TRBC1 antibody. It is considered that stimulation with a bispecific antibody that has the same antigen-binding portion as the anti-TRBC1 antibody used in this experiment would similarly increase the number of T cells that induce cytotoxicity.

[0170] Evaluation of T cell activation markers CD69 is a cell surface glycoprotein that is expressed very early after T or B cell activation, but not expressed on quiescent lymphocytes. Cytotoxic T cells have intracellular granules containing cytotoxic proteins (Granzyme B and Perforin). Granzyme B is a serine protease that, once delivered into a target cell, cleaves various intracellular substrates including caspases to induce apoptosis of the target cell. Perforin disrupts the plasma membrane of the target cell to facilitate the transport of Granzyme B into the target cell.

[0171] Fig. 6 shows the results of the analysis of the ratios of CD25, CD69, Granzyme B, and Perforin-positive cells in the T-LAK cells performed on Day 0 and Day 15.

[0172] In this experiment, the ratio of CD25-positive cells decreased, but the ratios of CD69, Granzyme B, and Perforin-positive cells increased, suggesting that TRBC1-positive cells were activated by the anti-TRBC1 antibody. The same effect is expected for a bispecific antibody having the same antigen-binding portion as the anti-TRBC1 antibody used in this experiment. It was estimated that, in this experiment, IL-2 was added during the culture, and therefore internalization or downregulation of CD25 occurred, resulting in decreased CD25 expression.

[0173] Change in ratio of PD-1-positive cells Fig. 7 shows the results of the analysis of the ratio of cells positive for PD-1, one of the immunosuppressive checkpoint molecules in T cells, in the T-LAK cells on Day 0 and Day 15. Expression of immunosuppressive checkpoint molecules such as PD-1 and TIGIT was scarcely observed in the effector memory (EM) cells of functional cytotoxic T cells, whereas high expression of such immunosuppressive checkpoint molecules as mentioned above is observed in non-functional exhausted T cells. As seen in the results shown in Fig. 7, the ratio of PD-1-positive cells did not change significantly. This suggested that T cells are less likely to be exhausted after T cell activation with anti-TRBC1 antibody. It is considered that T cells are similarly unlikely to be exhausted with a bispecific antibody having the same antigen-binding portion as the anti-TRBC1 antibody used in this experiment.

[0174] Quantification of released cytokines Fig. 8 shows the results of quantification of released amounts of IFN-y, IL-2, IL-6, IL -10, and TNFa of the T-LAK cells on Day 15 after stimulation with 0.1, 1.0, or 10 nM anti-TRBC1 antibody for 24 hours. As a result, antibody concentrationdependent increase in the released amount was observed for all the cytokines, and it was considered that the anti-TRBC1 antibody induced T cell activation and thus cytokines were released. In this experimental system, the released amount of IL-6, an inflammatory cytokine, was very small. These data indicate that the risk of cytokine release syndrome due to anti-TRBC1 antibody may be low. A low risk of cytokine release syndrome is similarly suggested for a bispecific antibody having the same antigen-binding portion as the anti-TRBC1 antibody used in this example.

[0175] <Reference Example 2> Evaluation of agonist activity of anti-TRBC2 antibody An anti-TRBC2 antibody was prepared and purified as described in Example 1. The heavy and light chain sequences of the anti-TRBC2 antibody are shown as SEQ ID NOs: 74 and 75. The reagents and antibodies for immunostaining used in this reference example are the same as those described in Reference Example 1.

[0176] Results Number of TRBC2-positive cells Human PBMCs were thawed and washed in a 37°C water bath. The washed human PBMCs were allowed to react with JOVI-1 mIgG (Santa Cruz) at 4°C for 30 minutes. PBMCs reacted with the antibody were washed, suspended in a medium, and allowed to react with Dynabeads M-280 Anti-Mouse IgG (Invitrogen) at 4°C for 30 minutes. The beads binding unwanted cells were collected on the tube wall by using DynaMag (Invitrogen), and the supernatant containing TRBC2-positive T cells was collected. The collected cells were cultured for 2 days in the RPMI 1640 medium containing 10% FBS supplemented with 10 nM anti-TRBC2 antibody and 100 IU / mL IL-2, then further cultured in the RPMI 1640 medium containing IL-2 and 10% FBS (up to 17 days), and thereby induced into T-LAK. The cells were subcultured every 2 or 3 days, and 100 IU / mL of IL-2 was newly added at the time of culture medium change.

[0177] Cell counts were measured at the start of the culture (Day 0) and on Day 15 by using an automated cell counter (BIO RAD, TC20). [Table 6] Culture# Priming Culture start date Cell Count Day0 (x106 cells) Cell Count Day15 (x106 cells) 1 hJovi-1-KFN December 12, 2022 6.4 19.1 2 hJovi-1-KFN December 12, 2022 7.07 18.6 3 hJovi-1-KFN January 18, 2023 5.68 42.3 The cell count, about 5 x 106 to 7 x 106 cells, on Day 0 increased to about 18 x 106 to 42 x 106 cells on Day 15 (Fig. 9). It was suggested that the addition of the anti-TRBC2 antibody selectively proliferated TRBC2-positive cells. It is considered that addition of a bispecific antibody having the same antigen-binding portion as the anti-TRBC2 antibody used in this experiment would similarly selectively proliferate TRBC2-positive cells.

[0178] Increase in cytotoxic T cells As shown in Fig. 10, the anti-TRBC2 antibody-sensitized T-LAK cells were stained with CD27 and CD45RA on Day 0 and Day 15, and the cell ratios of the effector memory (EM), central memory (CM), naive, and EMRA fractions were calculated. As a result, on Day 15, the ratio of the effector memory cells significantly increased to about 81% from about 27%, while the ratios of the other fractions decreased. These results suggest that stimulation with anti-TRBC2 antibody can increase the number of T cells that induce cytotoxicity. It is considered that stimulation with a bispecific antibody that has the same antigen-binding portion as the anti-TRBC2 antibody used in this experiment may similarly increase the number of T cells that induce cytotoxicity. The CD3-positive and TRBC1-negative T-LAK cell population was defined as TRBC2-positive cells.

[0179] Evaluation of T cell activation markers Fig. 11 shows the results of analysis of the ratios of CD25, CD69, Granzyme B, and Perforin-positive cells in the anti-TRBC2 antibody-sensitized T-LAK cells on Day 0 and Day 15. In this experiment, the ratio of CD25-positive cells decreased, while the ratios of CD69, Granzyme B, and Perforin-positive cells increased, suggesting that TRBC2-positive cells were activated by the anti-TRBC2 antibody. The same effect is expected for a bispecific antibody having the same antigen-binding portion as the anti-TRBC2 antibody used in this experiment. It is estimated that, in this experiment, IL-2 was added during the culture, and therefore internalization or downregulation of CD25 occurred, resulting in decreased CD25 expression. The CD3-positive and TRBC1-negative cell population was defined as TRBC2-positive cells.

[0180] Change in ratio of PD-1-positive cells Fig. 12 shows the results of analysis of the ratio of cells positive for PD-1, one of the immunosuppressive checkpoint molecules in T cells, in the anti-TRBC2 antibody-sensitized T-LAK cells on Day 0 and Day 15. Expression of immunosuppressive checkpoint molecules such as PD-1 and TIGIT was scarcely observed in the effector memory (EM) cells of functional cytotoxic T cells, whereas high expression of such immunosuppressive checkpoint molecules as mentioned above is observed in non-functional exhausted T cells. As seen in the results shown in Fig. 12, the ratio of PD-1-positive cells did not change significantly. This suggests that T cells are less likely to be exhausted when activated with anti-TRBC2 antibody, as in the case of anti-TRBC1 antibody. It is considered that T cells are similarly unlikely to be exhausted by a bispecific antibody having the same antigen-binding portion as that of the anti-TRBC2 antibody used in this experiment.

[0181] Quantification of released cytokines Fig. 13 shows the results of quantification of released amounts of IFN-y, IL-2, IL-6, IL -10, and TNFa of the T-LAK cells on Day 15 after stimulation with 0.1, 1.0, or 10 nM anti-TRBC2 antibody for 24 hours. As a result, antibody concentration dependent increase in the released amount was observed for all the cytokines, and it was considered that the anti-TRBC2 antibody induced T cell activation and thus the cytokines were released. In this experimental system, the released amount of IL-6, an inflammatory cytokine, was very small. These data indicate that the risk of cytokine release syndrome due to anti-TRBC2 antibody may be low. A low risk of cytokine release syndrome is similarly suggested for a bispecific antibody having the same antigen-binding portion as the anti-TRBC2 antibody used in this experiment.

[0182] <Example 3> Comparison of activities between the existing bispecific antibody and a TRBC1-based bispecific antibody In this example, to evaluate the in vitro efficacy of the bispecific antibodies prepared in Example 1, target cells were cultured with the T-LAK cells (effector cells) in the presence of the bispecific antibodies, and then cell viability of the target cells was measured by using a fluorescent dye (calcein activity) as an indicator to evaluate effector cell-dependent cytotoxicity. The experiment was performed with reference to Neri S et al. (Clin Diagn Lab Immunol. 2001 Nov;8(6):1131-5). Currently marketed bispecific T cell engagers mainly use anti-CD3 antibodies. In this example, the cytotoxicities of the anti-CD19 / anti-TRBC1 bispecific antibody (CD190001) and blinatumomab, an anti-CD19 / anti-CD3 bispecific antibody marketed as a therapeutic agent for B cell malignancies, were verified. It was suggested that bispecific antibodies prepared using anti-TRBC1 or anti-TRBC2 antibodies having an agonist activity can suppress cytokine release while maintaining sufficient cytotoxicity against tumors.

[0183] The following reagents were used HBSS (Hanks balanced salt solution), no phenol red (Gibco, 14025092) RPMI 1640 Medium, no phenol red (Gibco, 11835-030) Probenecid Water Soluble (Invitrogen, P36400) Calcein-AM (Invitrogen, C1430) Lysis Solution 10X (Promega, G1821)

[0184] Cytotoxicity of bispecific antibody against Raji cells Methods Bispecific antibodies adjusted to a final concentration of 1 fM to 100 nM (the concentration was varied depending on the antigen) with RPMI no phenol red medium (assay medium) containing 5 mM Probenecid and 0.5% FBS were added to a round bottom 96-well culture plate. A well containing only the assay medium was set as a control. For cytotoxicity assay using bispecific antibodies having a CD19-binding portion, Raji cells (target cells) were collected from the culture medium in a required amount and washed twice with HBSS. The cells were incubated at 37°C for 30 minutes in 10 ^M Calcein-AM supplemented with 5 mM Probenecid to label the target cells with calcein. After the incubation, the cells were washed three times with HBSS and suspended in the assay medium after the last washing. After counting the number of labeled target cells, the cells were added to the wells of the round-bottom 96-well culture plate in which bispecific antibodies were prepared. The effector cells were collected in a required amount from the culture medium, washed once with the assay medium, counted, and added to the wells of the round bottom 96-well culture plate in which the bispecific antibodies were prepared. The effector cells and the labeled target cells were added at a ratio of 10:1. The bispecific antibodies, labeled target cells, and effector cells on the plate were allowed to react at 5% CO2 and 37°C for 2 hours. After the round bottom 96-well culture plate was centrifuged (300 g, 5 minutes), the supernatants were collected, and the fluorescence of calcein eluted from the labeled target cells in the supernatants was measured by using a fluorescence plate reader (Bio Tek, SYNERGY H1). The cytotoxicity percentage of the target cells was calculated by using the fluorescence value of calcein as an index to evaluate the bispecific antibodydependent cytotoxicity.

[0185] The cytotoxicity percentage was determined as follows. The minimum eluted fluorescence value (lysis rate = 0%) of the labeled target cells was obtained from a well in which the labeled target cells and effector cells were incubated without any bispecific antibody. The maximum eluted fluorescence value (lysis rate = 100%) was obtained from a well containing the same content as that used for obtaining the minimum eluted fluorescence value and further containing 10x Lysis Solution. The cytotoxicity percentage was calculated in accordance with the following equation. Cytotoxicity percentage = [(Eluted fluorescence value of target cells in the presence of bispecific antibody) - (Minimum eluted fluorescence value)] / [(Maximum eluted fluorescence value) - (Minimum eluted fluorescence value)] S-shaped dose-response curves were calculated by using Prism Software (GraphPad Software Inc.), and EC50 values (50% effective concentration) and Emax values (maximum activity value) were calculated.

[0186] Results Fig. 14 and the following table show the results of the measurement of the cytotoxicities of the bispecific antibodies CD190001 and blinatumomab performed by adding the T-LAK cells to the Burkitt lymphoma-derived cell line expressing CD19, Raji cells. [Table 7] EC50 (pM) Standard deviation CD190001 1.9 0.4 Blinatumomab 0.12 0.02 Emax (%) Standard deviation CD190001 99.3 6.7 Blinatumomab 99.0 5.5 When T cells expressing CD3 and TRBC1 were present in sufficient amounts, both CD190001 and blinatumomab exhibited comparable activities, with Emax of approximately 99%.

[0187] <Example 4> Comparison of activities between the existing bispecific antibody and a TRBC1-based bispecific antibody The following reagents were used. PBS (Nacalai Tesque, 14249-24) Ficoll-PaqueTM PLUS (GE Healthcare, 17-1440-02) Leucosep (Grener Bio-One, 227290) FBS (NICHIREI, 175012, Lot. 19J00C) 2% FBS-PBS Red Blood Cell Lysis Buffer (Roche, 11814389001) CELLBUNKER2 (Takara, 11914) RPMI 1640 (Nacalai Tesque, 30264-85) 1 M Hepes (Nacalai, 17557-94) 100x Sodium pyruvate (Gibco, 11360-070) 100x MEM NEAA (Gibco, 11140-050) 100x 2-ME (Wako, 198-15781) PKH26 Red Fluorescent Cell Linker Kit (Sigma; PKH26GL) In this example, the cytotoxicities of the anti-CD19 / anti-TRBC1 bispecific antibody of the present invention and blinatumomab, which is marketed as a therapeutic agent for B cell malignancies, were verified. Blinatumomab is a bispecific antibody that targets CD19 and CD3 for use in the treatment of B cell malignancies. It is considered that use of human T cells in this example allows evaluation of cytotoxicity under conditions closer to those in vivo. The anti-CD19 / anti-CD3 bispecific antibody was produced as described in Example 1.

[0188] Methods For the cytotoxicity assay using bispecific antibodies with antigen-binding portions targeting CD19 and either TRBC1 or CD3, CD190001 and blinatumomab prepared in Example 3 were used. The bispecific antibodies were adjusted to a final concentration of 1 fM to 1 nM by using the RPMI 1640 medium containing 5% FBS, 10 mM HEPES, 1 mM sodium pyruvate, MEM NEAA, and 0.1 mM 2-ME (TDCC medium), and added to a U-bottom 96-well culture plate. As a control, a well containing only the TDCC medium was prepared. Raji cells (target cells) in a required amount were collected from the culture medium, labeled with PKH26 according to the protocol of PKH26 Red Fluorescent Cell Linker Kit, and suspended in the TDCC medium, and the cell number was counted. The cells were then added to wells of the U-bottom 96-well culture plate in which the bispecific antibodies were prepared. PBMCs derived from healthy individuals (effector cells) were collected in a required amount, and suspended in the TDCC medium, and the cell number was counted. The cells were then added to the wells of a flat bottom 96-well plate in which the bispecific antibodies were prepared. The effector cells and the labeled target cells were added at a ratio of 20:1. The plate containing the bispecific antibodies, labeled target cells, and effector cells was incubated at 5% CO2 and 37°C. After 72 hours, the cells were transferred from each well to a V-bottom 96-well plate, and washed once with 2% FBS / PBS, then dead cells were stained, and a diluted solution of antibody for T cell activation marker was added. After incubation at 4°C for 30 minutes, the cells were washed twice with 2% FBS / PBS, CountBright Absolute Counting Beads (Invitrogen, C36950) were added to the samples, and measurement was performed using a flow cytometer (BD Symphony A1). The PKH26-positive and Fixable Viability Dye eFluor780-negative fraction was gated as live target cells, cell count was corrected using CountBright Absolute Counting Beads, and cytotoxic activity was evaluated. FlowJo v10.9.0 was used as the analysis software.

[0189] The cytotoxicity percentage was determined as follows. For the live target cell count, the number of cells in the PKH26-positive and Fixable Viability Dye eFluor780-negative fraction was corrected to the number of cells per 3,000 CountBright Absolute Counting Beads. The maximum live target cell count (cytotoxicity = 0%) was obtained from a well in which the labeled target cells and effector cells were incubated without any bispecific antibody. The cytotoxicity percentage was calculated in accordance with the following equation. Cytotoxicity percentage = 100 - [100 x (Live target cell count in the presence of bispecific antibody) / (Maximum live cell count)] S-shaped dose-response curves were obtained by calculation using Prism Software (GraphPad Software Inc.), and EC50 value (50% effective concentration) and Emax value (maximum activity value) were calculated. Preparation of healthy human-derived PBMCs PBMCs were prepared from blood collected from healthy subjects as follows. Blood heparinized at the time of blood collection was diluted 3-fold with PBS, added to a Leucosep tube filled with Ficoll and diluted with PBS to 50 mL. The tube was then centrifuged (1000 x g, 10 minutes, 20°C), the upper layer was removed, and the PBMC layer was transferred to a separate tube. The cells were washed twice with PBS, and suspended in 3 mL of 2% FBS-PBS, 6 mL of RBC Buffer was added to the suspension, and the mixture was allowed to stand for 10 minutes. The cells were washed with 10 mL of 2% FBS / PBS, suspended in CELLBUNKER2, and stored in liquid nitrogen until use.

[0190] Results Fig. 15 shows the results of measuring the cytotoxicities of the anti-CD19 / anti-TRBC1 bispecific antibody (CD190001) and the anti-CD19 / anti-CD3 bispecific antibody (blinatumomab) against Raji cells, a cell line derived from Burkitt lymphoma, observed when PBMCs derived from healthy individuals were added to the Raji cells. The EC50 values and Emax values (N=4) are shown in the following table. [Table 8] EC50 (pM) Standard deviation CD190001 1.2 0.7 Blinatumomab 0.04 0.02 Emax (%) Standard deviation CD190001 76.8 1.7 Blinatumomab 90.7 3.0 Both the anti-CD19 / anti-TRBC1 bispecific antibody and anti-CD19 / anti-CD3 bispecific antibody exhibited sufficient cytotoxicity. Although approximately 50% of T cells are TRBC1-positive, the activity of the bispecific antibody targeting TRBC1 is believed to be comparable to that of the bispecific antibody targeting CD3.

[0191] <Example 5> Changes in T cell activation markers induced by the existing bispecific antibody and a TRBC1-based bispecific antibody In this example, it was demonstrated that a TRBC1-based bispecific antibody resulted in a lower proportion of T cells expressing activation markers, compared with a CD3-based bispecific antibody. This suggests that TRBC1- or TRBC2-based bispecific antibodies can reduce the risk of cytokine release syndrome etc. by selectively activating only a subset of T cells.

[0192] Methods To evaluate T cell activation by bispecific antibodies having antigen-binding portions for CD19 and either TRBC1 or CD3, the antibodies used in Example 4 were employed. The bispecific antibodies were each adjusted to a final concentration of 1 fM to 1 nM in RPMI 1640 medium containing 5% FBS, 10 mM HEPES, 1 mM sodium pyruvate, MEM NEAA, and 0.1 mM 2-ME (TDCC medium), and added to a U-bottom 96-well culture plate. A well containing only the TDCC medium was set as a control. Raji cells (target cells) were collected in a required amount from the culture medium, labeled with PKH26 according to the protocol of PKH26 Red Fluorescent Cell Linker Kit, and suspended in the TDCC medium, and the cell number was counted. The cells were then added to the U-bottom 96-well culture plate in which the bispecific antibodies were prepared. PBMCs (effector cells) derived from healthy individuals were collected in a required amount, and suspended in the TDCC medium, and the cell number was counted. The cells were then added to a flat-bottom 96-well culture plate in which the bispecific antibodies were prepared. The effector cells and labeled target cells were added at a ratio of 20:1. The plate containing the bispecific antibodies, labeled target cells, and effector cells was incubated at 37°C in a 5% CO2 atmosphere. After 72 hours, the cells from each well were transferred to a V-bottom 96-well plate, and washed once with 2% FBS / PBS, then dead cells were stained, and a diluted solution of antibody for T cell activation marker was added. After incubation at 4°C for 30 minutes, the cells were washed twice with 2% FBS / PBS, CountBright Absolute Counting Beads (Invitrogen, C36950) were added to the samples, and measurement was performed using a flow cytometer (BD Symphony A1). The PKH26-negative and Fixable Viability Dye eFluor 780-negative fraction was gated as viable effector cells, and rates of cells positive for CD25 and CD69 as activation markers for CD4- or CD8positive T cells were evaluated. [Table 9] Antibody Dye-labeled Manufacturer Item number Anti-CD4 antibody BV650 BioLegend 317435 Anti-CD8 antibody APC BioLegend 344721 Anti-CD25 antibody BV421 BioLegend 302630 Anti-CD69 antibody PerCP-Cy5.5 BioLegend 310925 *BV421, BV650: Abbreviations for Brilliant Violet 421 (registered trademark) and Brilliant Violet 650 (registered trademark)

[0193] Results The results are shown in Fig. 16. When the expression of CD25 and CD69, which are activation markers for CD4- or CD8-positive T cells, was measured in the effector cells following their reaction with the target cells and the bispecific antibody used in this example, both markers were induced in an antibody concentrationdependent manner. When CD190001 was compared with blinatumomab, the proportion of activated T cells was 50% or less in the presence of the antibodies at the same concentration. These results suggest that, with the TRBC1-based bispecific antibody, only T cells expressing TRBC1 are cross-linked with the target cells and function as effector cells.

[0194] <Example 6> Comparison of the existing bispecific antibody and a TRBC1-based bispecific antibody for cytokine release The following reagents were used. PBS (Nacalai Tesque, 14249-24) Ficoll-PaqueTM PLUS (GE Healthcare, 17-1440-02) Leucosep (Greiner Bio-One, 227290) FBS (Nichirei, 175012, Lot 19J00C) 2% FBS-PBS Red Blood Cell Lysis Buffer (Roche, 11814389001) CELLBUNKER2 (Takara, 11914) RPMI 1640 (Nacalai Tesque 30264-85) 1 M Hepes (Nacalai, 17557-94) 100x Sodium pyruvate (Gibco, 11360-070) 100x MEM NEAA (Gibco, 11140-050) 100x 2-ME (Wako, 198-15781) PKH26 Red Fluorescent Cell Linker Kit (Sigma, PKH26GL) MILLIPLEX (registered trademark) Human Cytokine / Chemokine / Growth Factor Panel A (Millipore, HCYTA-60K)

[0195] Methods To evaluate activation of T cells by bispecific antibodies having antigenbinding portions for CD19 and either TRBC1 or CD3, the antibodies used in Example 4 were employed. The bispecific antibodies were each adjusted to a final concentration of 1 fM to 1 nM in RPMI 1640 medium containing 5% FBS, 10 mM HEPES, 1 mM sodium pyruvate, MEM NEAA, and 0.1 mM 2-ME (TDCC medium), and added to a U-bottom 96-well culture plate. A well containing only the TDCC medium was set as a control. Raji cells (target cells) were collected in a required amount from the culture medium, and suspended in the TDCC medium, and the cell number was counted. The cells were then added to a U-bottom 96-well culture plate in which the bispecific antibodies were prepared. PBMCs (effector cells) derived from healthy individuals were collected in a required amount, and suspended in the TDCC medium, and the cell number was counted. The cells were then added to the flat-bottom 96-well culture plate in which the bispecific antibodies were prepared. The effector cells and labeled target cells were added at a ratio of 20:1. The plate containing the bispecific antibodies, labeled target cells, and effector cells was incubated at 37°C in a 5% CO2 atmosphere. After 24 hours, the culture supernatant was collected and stored in a -80°C freezer until measurement. Samples were prepared according to the protocol for MILLIPLEX (registered trademark) Human Cytokine / Chemokine / Growth Factor Panel A, and IFN-y, IL-2, and TNF-a were quantified. Quantification values were automatically calculated on the Luminex (registered trademark) 200xPONET (registered trademark) 3.1 system.

[0196] Results The results are shown in Fig. 17. In comparison for the major cytokines, IFN- Y, TNF-a, and IL-2, it was found that both blinatumomab and CD190001 caused an increase in cytokine release in an antibody concentration-dependent manner. In comparison for the maximum release levels within the measurement range, the cytokine levels observed with CD190001 were less than half those observed with blinatumomab for all the cytokines. It is thought that the proportion of TRBC1-positive cells among T cells is approximately 50%, thus the number of T cells that are activated as effector cells and release cytokines is also small, and therefore the total amount of cytokine release can be suppressed.

[0197] <Example 7> Cytotoxicity of an anti-CD19 / anti-TRBC1 bispecific antibody In this example, to evaluate the in vitro efficacy of the bispecific antibodies prepared in Example 1, target cells were cultured with T-LAK cells (effector cells) in the presence of each bispecific antibody, and then the cell viability of the target cells was measured by using a fluorescent dye (calcein activity) as an indicator to evaluate effector cell-dependent cytotoxicity. The experiment was conducted with reference to Neri S et al. (Clin Diagn Lab Immunol. 2001 Nov;8(6):1131-5). The following reagents were used: HBSS (Hank’s Balanced Salt Solution), no phenol red (Gibco, 14025092) RPMI 1640 Medium, no phenol red (Gibco, 11835-030) Probenecid Water-Soluble (Invitrogen, P36400) Calcein-AM (Invitrogen, C1430) Lysis Solution 10x (Promega, G1821)

[0198] Cytotoxicity of bispecific antibodies against Raji cells Methods The bispecific antibodies were each adjusted to a final concentration of 1 fM to 100 nM (concentration varied depending on the antigen) in the RPMI 1640, no phenol red medium (assay medium) containing 5 mM probenecid and 0.5% FBS, and added to a round-bottom 96-well culture plate. A well containing only the assay medium was set as a control. For the cytotoxicity assay using a bispecific antibody having a CD19-binding portion, Raji cells (target cells) were collected in a required amount from the culture medium and washed twice with HBSS. The target cells were labeled with calcein by incubating them at 37°C for 30 minutes with 10 ^M Calcein-AM containing 5 mM probenecid. After the incubation, the cells were washed three times with HBSS, and, following the final wash, suspended in the assay medium. The labeled target cells were counted, and then added to the round-bottom 96-well culture plate in which the bispecific antibodies were prepared. The effector cells were collected in a required amount from the culture medium, washed once with the assay medium, counted, and then added to the round-bottom 96-well culture plate in which the bispecific antibodies were prepared. The effector cells and labeled target cells were added at a ratio of 10:1. The plate containing the bispecific antibodies, labeled target cells, and effector cells was incubated at 37°C in a 5% CO2 atmosphere for 2 hours. After centrifuging the roundbottom 96-well culture plate (300 g, 5 minutes), the supernatant was collected, and the fluorescence of calcein eluted from the labeled target cells in the supernatant was measured using a plate reader (Bio-Tek, SYNERGY H1). The cytotoxicity percentage of the target cells was calculated using the calcein fluorescence values as an indicator to evaluate bispecific antibody-dependent cytotoxicity.

[0199] The cytotoxicity percentage was determined as follows. The minimum eluted fluorescence value (lysis rate = 0%) of the labeled target cells was obtained from a well in which the labeled target cells and effector cells were incubated without any bispecific antibody. The maximum eluted fluorescence value (lysis rate = 100%) was obtained from a well containing the same content as that used for obtaining the minimum eluted fluorescence value and further containing 10x Lysis Solution. The cytotoxicity percentage was calculated in accordance with the following equation. Cytotoxicity percentage = [(Eluted fluorescence value of target cells in the presence of bispecific antibody) - (Minimum eluted fluorescence value)] / [(Maximum eluted fluorescence value) - (Minimum eluted fluorescence value)] S-shaped dose-response curves were calculated by using Prism Software (GraphPad Software Inc.), and EC50 values (50% effective concentration) and Emax values (maximum activity value) were calculated.

[0200] Results Fig. 18 and the following table show the results of measuring the cytotoxicity provided by the bispecific antibody CD190001 against the Raji cells, a cell line derived from Burkitt lymphoma that expresses CD19, to which T-LAK cells were added. The EC50 and Emax values (N=3) are shown in the following table. [Table 10] EC50 (pM) Standard deviation CD190001 1.4 0.5 Emax (%) Standard deviation CD190001 97.1 6.8 The anti-CD19 / anti-TRBC1 bispecific antibody CD190001 was expected to exhibit extremely potent cytotoxicity against CD19-expressing cells.

[0201] <Example 8> Cytotoxicity of an anti-BCMA / anti-TRBC1 bispecific antibody In this example, to evaluate the in vitro efficacy of the anti-BCMA / anti-TRBC1 bispecific antibody prepared in Example 1, target cells U266 were cultured with anti-TRBC1 antibody-sensitized T-LAK (effector cells) in the presence of the bispecific antibody, and then the cell viability of the target cells was measured using a fluorescent dye (calcein activity) as an indicator to evaluate the effector cell-dependent cytotoxicity. The target cells, U266, are a cell line that expresses BCMA. Cytotoxicity was measured using the same method as described in Example 3.

[0202] Results Fig. 19 and the following table show the results of measuring the cytotoxicity provided by the bispecific antibody BCMA0001 against U266 cells, a multiple myeloma-derived cell line that expresses BCMA, to which the T-LAK cells were added. The EC50 and Emax values (N=3) are shown in the following table. [Table 11] Antibody Name EC50 (pM) Standard deviation BCMA0001 159.2 62.1 Antibody Name Emax (%) Standard deviation BCMA0001 97.8 7.3 The anti-BCMA / anti-TRBC1 bispecific antibody BCMA0001 was expected to exhibit extremely potent cytotoxicity against BCMA-expressing cells.

[0203] <Example 9> Cytotoxicity of anti-CD33 / anti-TRBC1 bispecific antibodies In this example, to evaluate the in vitro efficacy of the anti-CD33 / anti-TRBC1 bispecific antibodies prepared in Example 1, target cells MOLM13 were cultured with anti-TRBC1 antibody-sensitized T-LAK cells (effector cells) in the presence of each bispecific antibody, and the cell viability of the target cells was then measured using a fluorescent dye (calcein activity) as an indicator to evaluate effector cell-dependent cytotoxicity. The target cells, MOLM13, are a cell line that expresses CD33. Cytotoxicity was measured using the same method as described in Example 3.

[0204] Results Fig. 20 and the following table show the results of measuring the cytotoxicity provided by the bispecific antibody against MOLM13 cells, a cell line derived from acute myeloid leukemia that expresses CD33, to which the T-LAK cells were added. The EC50 and Emax values (N=3) are shown in the following table. [Table 12] Antibody Name EC50 (pM) Standard deviation CD330001 349.8 76.0 CD330003 33.4 4.5 Antibody Name Emax (%) Standard deviation CD330001 88.3 5.7 CD330003 75.8 0.7 The anti-CD33 / anti-TRBC1 bispecific antibodies were expected to exhibit extremely potent cytotoxicity against cells expressing the antigen.

[0205] <Example 10> Cytotoxicity of anti-CD123 / anti-TRBC1 bispecific antibodies In this example, to evaluate the in vitro efficacy of the anti-CD123 / anti-TRBC1 bispecific antibodies prepared in Example 1, target cells MOLM13 were cultured with anti-TRBC1 antibody-sensitized T-LAK cells (effector cells) in the presence of each bispecific antibody, and the cell viability of the target cells was then measured using a fluorescent dye (calcein activity) as an indicator to evaluate effector cell-dependent cytotoxicity. The target cells, MOLM13, are a cell line that expresses CD123. Cytotoxicity was measured using the same method as described in Example 3.

[0206] Results Fig. 21 and the following table show the results of measuring the cytotoxicity provided by each bispecific antibody against MOLM13 cells, a CD123-expressing cell line derived from acute myeloid leukemia, to which T-LAK cells were added. The EC50 and Emax values (N=3) are shown in the following table. [Table 13] Antibody Name EC50 (pM) Standard deviation CD1230001 15.5 3.5 CD1230003 14.9 5.5 Antibody Name Emax (%) Standard deviation CD1230001 64.0 6.9 CD1230003 61.2 7.2 The anti-CD123 / anti-TRBC1 bispecific antibodies were expected to exhibit extremely potent cytotoxicity against cells expressing the antigen.

[0207] <Example 11> Cytotoxicity of an anti-HER2 / anti-TRBC1 bispecific antibody In this example, to evaluate the in vitro efficacy of the anti-HER2 / anti-TRBC1 bispecific antibody prepared in Example 1, target cells MKN7 were cultured with anti-TRBC1 antibody-sensitized T-LAK cells (effector cells) in the presence of the bispecific antibody, and the cell viability of the target cells was then measured using a fluorescent dye (calcein activity) as an index to evaluate effector cell-dependent cytotoxicity. Methods The target cells, MKN7, are a cell line that expresses HER2. Two days prior to the activity assay, the target cells were seeded on a 96-well culture plate and incubated at 37°C in a 5% CO2 atmosphere. On the day of the assay, the plate on which the target cells were seeded was washed three times with HBSS, and the target cells were labeled with calcein by incubating them with 10 pM Calcein-AM containing 5 mM probenecid at 37°C for 30 minutes. After the incubation, the plate was washed three times with HBSS, and following the final wash, the bispecific antibody diluted to each concentration was added to the 96-well culture plate. The effector cells were collected in a required amount from the culture medium, washed once with the assay medium, counted, and then added to the 96-well culture plate. The effector cells and labeled target cells were added at a ratio of 10:1. The plate containing the bispecific antibody, labeled target cells, and effector cells was incubated at 37°C in a 5% CO2 atmosphere for 2 hours. The 96-well culture plate was centrifuged (300 g, 5 minutes), then the supernatant was collected, and the fluorescence of calcein eluted from the labeled target cells in the supernatant was measured using a plate reader (Bio-Tek, SYNERGY H1). The cytotoxicity percentage against the target cells was calculated using the calcein fluorescence values as an indicator to evaluate the bispecific antibodydependent cytotoxicity. The cytotoxicity was calculated using the same method as described in Example 3.

[0208] Results Fig. 22 and the following table show the results of measuring the cytotoxicity provided by the bispecific antibody against the MKN7 cells, a cell line derived from HER2-expressing tubular adenocarcinoma, to which the T-LAK cells were added. The EC50 and Emax values (N=3) are shown in the following table. [Table 14] Antibody Name EC50 (pM) Standard deviation HER20001 20.6 2.5 Antibody Name Emax (%) Standard deviation HER20001 76.0 3.6 The anti-HER2 / anti-TRBC1 bispecific antibody was thought to exhibit extremely potent cytotoxicity against cells expressing the antigen. The anti-HER2 / anti-TRBC1 bispecific antibody was thought to exhibit effective cytotoxicity against solid tumors expressing the antigen.

Claims

1. A therapeutic agent for a malignancy (provided that the malignancy is not a T cell malignancy), comprising a bispecific antigen-binding molecule, wherein the bispecific antigen-binding molecule comprises(1) at least one portion that specifically binds to a target tumor antigen expressed on tumor cells, and(2) at least one portion that specifically binds to a normal T cell-side target antigen having subtypes, provided thatthe target tumor antigen expressed on the tumor cells is not present on normal T cells, or even if it is present, the normal T cells are not substantially activated when the bispecific antigen-binding molecule binds to the same antigen as the target tumor antigen present on the normal T cells, binding of the bispecific antigen-binding molecule to the normal T cell-side target antigen activates the normal T cells, anda sufficient proportion of a subtype of the normal T cell-side target antigen is present to provide a sufficient number of activated T cells for the treatment of the malignancy.

2. The therapeutic agent according to claim 1, wherein the normal T cell-side target antigen having subtypes is a TRBC (T cell receptor beta-chain constant region), andthe subtype of the normal T cell-side target antigen is TRBC1 or TRBC2.

3. The therapeutic agent according to claim 1 or 2, wherein the subtype of the normal T cell-side target antigen is TRBC1, the bispecific antigen-binding molecule comprises at least one portion that specifically binds to TRBC1 of the normal T cells, and the at least one portion comprises a VH domain comprisinga heavy chain CDR1 comprising the amino acid sequence of SEQ ID NO: 5,a heavy chain CDR2 comprising the amino acid sequence of SEQ ID NO: 6, anda heavy chain CDR3 comprising the amino acid sequence of SEQ ID NO: 7; anda VL domain comprisinga light chain CDR1 comprising the amino acid sequence of SEQ ID NO: 8, a light chain CDR2 comprising the amino acid sequence of SEQ ID NO: 9, anda light chain CDR3 comprising the amino acid sequence of SEQ ID NO: 10.

4. The therapeutic agent according to claim 3, wherein the at least one portion that specifically binds to TRBC1 of the normal T cells comprises a heavy chain variable region VH that is at least 90% identical to the amino acid sequence of SEQ ID NO: 1 and a light chain variable region VL that is at least 90% identical to the amino acid sequence of SEQ ID NO: 2, or a heavy chain variable region VH that is at least 90% identical to the amino acid sequence of SEQ ID NO: 90 and a light chain variable region VL that is at least 90% identical to the amino acid sequence of SEQ ID NO: 91.

5. The therapeutic agent according to claim 1 or 2, wherein the subtype of the normal T cell-side target antigen is TRBC2, the bispecific antigen-binding molecule comprises at least one portion that specifically binds to TRBC2 of the normal T cells, and the at least one portion comprises a VH domain comprisinga heavy chain CDR1 comprising the amino acid sequence of SEQ ID NO: 11, a heavy chain CDR2 comprising the amino acid sequence of SEQ ID NO: 12, and a heavy chain CDR3 comprising the amino acid sequence of SEQ ID NO: 13; and a VL domain comprisinga light chain CDR1 comprising the amino acid sequence of SEQ ID NO: 14,a light chain CDR2 comprising the amino acid sequence of SEQ ID NO: 15, anda light chain CDR3 comprising the amino acid sequence of SEQ ID NO: 16.

6. The therapeutic agent according to claim 5, wherein the at least one portion that specifically binds to TRBC2 of the normal T cells comprises a heavy chain variable region VH that is at least 90% identical to the amino acid sequence of SEQ ID NO: 3 and a light chain variable region VL that is at least 90% identical to the amino acid sequence of SEQ ID NO: 4.

7. The therapeutic agent according to claim 1 or 2, wherein the malignancy is a hematological malignancy other than T cell malignancy.

8. The therapeutic agent according to claim 7, wherein the hematological malignancy is a myeloid malignancy or B cell malignancy.

9. The therapeutic agent according to claim 1 or 2, wherein the malignancy is asolid tumor.

Images