Anti-CD26 antibody and its applications

Abstract

The present invention provides an antibody or antigen-binding fragment that binds to human CD26, and uses the antibody or antigen-binding fragment in the manufacture of a drug for treating tumors with high CD26 expression. [Solution] The present invention provides a CD26 mouse-derived antibody and a chimeric or humanized antibody derived from the same as a parent antibody, which realizes superior antigen affinity and good safety and efficacy compared to other mouse antibodies and corresponding chimeric or humanized antibodies. The present invention also provides a bispecific T cell-inducing antibody targeting CD26, which has better safety and efficacy compared to existing full-length CD26 antibodies, and realizes reduced dosage, better therapeutic effect, and reduced incidence of adverse reactions in disease treatment. Furthermore, the bispecific T cell-inducing antibody targeting CD26 of the present invention exhibits outstanding efficacy and safety compared to existing first-line treatments for related tumors (e.g., treatment of kidney cancer with BAVENCIO).

Description

[Technical Field]

[0001] The present invention relates to antibodies or antigen-binding fragments that bind to human CD26, and the use of these antibodies or antigen-binding fragments in the manufacture of drugs for treating tumors with high CD26 expression. [Background technology]

[0002] CD26 is a multifunctional type II transmembrane glycoprotein that is also present in soluble form in plasma. CD26 frequently exists as a homodimer, with the monomer containing 766 amino acids and a relative molecular mass of approximately 110 kDa. The amino acid residues of CD26 are divided into five parts: the intracellular domain (1–6), the transmembrane domain (7–28), the highly glycosylated domain (29–323), the cysteine-rich domain (324–551), and the C-terminal catalytic domain (552–766). The three-dimensional structure of the molecule is closely related to its function. CD26 (DPP4) inhibitors have been used clinically for decades to treat type II diabetes. CD26 expression levels are significantly elevated on the surface of various tumor cells, including malignant mesothelioma, renal carcinoma, prostate cancer, and lung cancer. For these tumors with high CD26 expression, CD26 is considered to be an important target of action (CD26 / DPP4-a potential biomarker and target for cancer therapy, Pharmacology & Therapeutics(2019), 198: 135-159).

[0003] Currently, the monoclonal antibody YS110 from Y'S Therapeutics has made the most progress in research into targeted anti-cancer drugs targeting CD26, and has already completed Phase I / II clinical trials. However, analysis of existing literature indicates that YS110 has low activity both in vivo and in vitro in preclinical studies. For example, the paper "A humanized anti-CD26 monoclonal antibody inhibits cell growth of malignant mesothelioma via retarded G2 / M cell cycle transition, Cancer Cell Int (2016) 16:35" reports that after 48 hours of action of YS110 at a concentration of 250 μg / ml, the growth inhibition rate of tumor cell lines was 18.3%, while the IC 50 The values ​​are much higher than 250 μg / ml, suggesting relatively low in vitro activity. Furthermore, the patent "Anti-CD26 Antibody and Method for Use Thereof (Application Number: CN200680034937.4)" demonstrated the treatment of mouse tumor models with various CD26-high-expressing cell lines with YS110. While YS110 reduced tumor size and inhibited tumor growth to some extent, it was unable to completely suppress tumor growth and achieve tumor cure. This indicates that its activity in animal models is still unsatisfactory. The results of a completed Phase II clinical trial of YS110 showed that although YS110 was well tolerated, its disease control rate was not as expected. This result corresponds to the low activity of YS110 demonstrated in preclinical studies.

[0004] Although literature and materials state that CD26 is widely distributed, the use of YS110 has been proven to have a favorable safety profile, which suggests that its safety is fully guaranteed for clinical use. Of the 31 patients who participated in clinical trials of YS110, no deaths occurred, and the main side effects were infusion reactions (16.1%), hiccups (9.7%), and diarrhea (6.5%). (Phase 2 Study of YS110, a Recombinant Humanized Anti-CD26 Monoclonal Antibody, in Japanese Patients With Advanced Malignant Pleural Mesothelioma, JTO Clinical and Research Reports (2021), 2(6)) [Prior art documents] [Patent documents]

[0005] [Patent Document 1] CN101282994B [Non-patent literature]

[0006] [Non-Patent Document 1] CD26 / DPP4-a potential biomarker and target for cancer therapy, Pharmacology & Therapeutics(2019), 198: 135-159 [Non-patent document 2] A humanized anti-CD26 monoclonal antibody inhibits cell growth of malignant mesothelioma via retarded G2 / M cell cycle transition, Cancer Cell Int(2016) 16:35 [Non-patent document 3] Phase 2 Study of YS110, a Recombinant Humanized Anti-CD26 Monoclonal Antibody, in Japanese Patients With Advanced Malignant Pleural Mesothelioma, JTO Clinical and Research Reports(2021), 2(6) Summary of the Invention [Problem to be solved by the invention]

[0007] As mentioned above, the safety of targeting CD26 has already been confirmed in clinical trials, but it is difficult to achieve sufficient therapeutic effects against tumors with existing antibodies that target CD26 therapeutically. [Means for solving the problem]

[0008] In an effort to solve the above problems, a first object of the present invention is to provide a more effective and safer CD26 antibody or antigen-binding fragment.

[0009] The antibody or antigen-binding fragment provided by the present invention that specifically binds to human CD26 comprises an HCDR1 shown in SEQ ID NO:1, an HCDR2 shown in SEQ ID NO:2, an HCDR3 shown in SEQ ID NO:3, and an LCDR1 shown in SEQ ID NO:4, an LCDR2 consisting of Arg Met Ser, and an LCDR3 shown in SEQ ID NO:5.

[0010] The antibody or antigen-binding fragment comprises a heavy chain variable region whose amino acid sequence is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to the sequence set forth in SEQ ID NO:6 or SEQ ID NO:8, and a light chain variable region whose amino acid sequence is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to the sequence set forth in SEQ ID NO:7 or SEQ ID NO:9.

[0011] The antibody or antigen-binding fragment may have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid insertions, deletions, or substitutions in the sequence set forth in SEQ ID NO:6, SEQ ID NO:7, SEQ ID NO:8, or SEQ ID NO:9. The amino acid substitutions are conservative amino acid substitutions.

[0012] The antibody or antigen-binding fragment comprises a heavy chain variable region set forth in SEQ ID NO:6 and a light chain variable region set forth in SEQ ID NO:7, or a heavy chain variable region set forth in SEQ ID NO:8 and a light chain variable region set forth in SEQ ID NO:9.

[0013] A second object of the present invention is to provide bispecific T cell-inducing antibodies (BiTEs) comprising the sequence of a CD26 antibody or antigen-binding fragment.

[0014] A bispecific T cell-inducing antibody (BiTE) comprising the sequence of a CD26 antibody or antigen-binding fragment of the present invention comprises two antigen-specific single-chain variable fragments (scFv) linked by a linker, one of which recognizes CD26 as a target and is connected by a linker between the antibody heavy chain variable region and antibody light chain variable region that recognize CD26, and the other scFv recognizes CD3 as a target and is connected by a linker between the antibody heavy chain variable region and antibody light chain variable region that recognize CD3.

[0015] The scFv that recognizes CD26 as a target comprises an HCDR1 shown in SEQ ID NO:1, an HCDR2 shown in SEQ ID NO:2, an HCDR3 shown in SEQ ID NO:3, an LCDR1 shown in SEQ ID NO:4, an LCDR2 consisting of Arg Met Ser, and an LCDR3 shown in SEQ ID NO:5.

[0016] The scFv that recognizes CD26 as a target comprises a heavy chain variable region whose amino acid sequence is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to the sequence set forth in SEQ ID NO:6 or SEQ ID NO:8, and a light chain variable region whose amino acid sequence is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to the sequence set forth in SEQ ID NO:7 or SEQ ID NO:9.

[0017] The scFv that recognizes CD26 as a target may have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid insertions, deletions, or substitutions in the sequence shown in SEQ ID NO: 6, SEQ ID NO: 7, SEQ ID NO: 8, or SEQ ID NO: 9. The amino acid substitutions are conservative. The scFv that recognizes CD26 as a target comprises a heavy chain variable region shown in SEQ ID NO:6 and a light chain variable region shown in SEQ ID NO:7, or a heavy chain variable region shown in SEQ ID NO:8 and a light chain variable region shown in SEQ ID NO:9.

[0018] The scFv that recognizes CD3 as a target is derived from the sequence of a CD3 antibody known in the art, such as OKT-3, L2K, TR66, UCHT1, SP34, IORT3, Catumaxomab, Blinatumomab, or Solitomab, for example, the anti-CD3 heavy chain variable region shown in SEQ ID NO: 10 and the anti-CD3 light chain variable region shown in SEQ ID NO: 11.

[0019] The linker connecting the heavy chain variable region and the light chain variable region of the scFv may be a linker commonly used in the art, such as KESGSVSSEQLAQFRSLD, EGKSSGSGSESKST, GSTSGGGSGGGSGGGGSS, GSTSGSGKPGSGEGSTKG, (GGGGS) nIn this, n may be an integer of 1 to 5, preferably an integer of 1 to 3, and more preferably, n may be 3 as shown in SEQ ID NO:12.

[0020] The linker connecting the two scFvs is a linker commonly used in the art, such as (GGGGS) n In this, n may be an integer of 1 to 5, preferably n may be an integer of 1 to 3, and more preferably n may be 1 as shown in SEQ ID NO:13.

[0021] The bispecific T cell-inducing antibody comprising the sequence of the CD26 antibody or antigen-binding fragment preferably has the amino acid sequence shown in SEQ ID NO:14 (amino acid sequence of 18G272 BiTE) or SEQ ID NO:15 (amino acid sequence of 19G294 BiTE).

[0022] The present invention also relates to a nucleotide sequence encoding the amino acid sequence of the antibody or antigen-binding fragment. Furthermore, the present invention relates to a vector containing the nucleotide sequence and a host cell containing the vector.

[0023] The present invention also relates to methods for producing antibodies or antigen-binding fragments, said methods comprising culturing said host cells and recovering the antibody or antigen-binding fragment from the culture.

[0024] The present invention also relates to a pharmaceutical composition comprising said antibody or antigen-binding fragment, which may further comprise a pharmaceutically acceptable excipient.

[0025] The present invention also relates to a method for treating tumors, which comprises administering to a patient an effective amount of the antibody or antigen-binding fragment. The tumor is a tumor that highly expresses CD26, including, but not limited to, renal cancer, mesothelioma, lung cancer, liver cancer, prostate cancer, etc. The antibody or antigen-binding fragment is administered alone or in combination with other anti-cancer agents. [Effects of the Invention]

[0026] The present invention has the following advantages over existing technologies: First, the murine-derived antibodies screened by the present invention are clearly superior. The data in Example 1 show that the murine antibody 62 screened by the present invention has a much higher ability to lyse human tumor cells than other murine antibodies. Furthermore, the murine antibody screened by the present invention has the same ability to kill human tumor cells as the humanized antibody YS110, which has completed a phase II clinical trial, further demonstrating the superiority of the murine antibody of the present invention.

[0027] Furthermore, the data in Examples 5, 8, and 10 show that chimeric or humanized antibodies engineered based on the variable region of murine antibody No. 62 have higher dot-blot activity and better antigen affinity than corresponding antibodies derived from other murine antibodies.

[0028] Compared with BiTEs derived from other murine antibodies, BiTEs engineered based on the variable region of the murine antibody No. 62 of the present invention exhibited higher efficacy. Example 11 shows the results of an in vitro comparison of the cytotoxicity of PBMCs mediated by BiTEs against different target cells, demonstrating that BiTEs derived from No. 62 of the present invention have superior killing ability against multiple tumor cells compared to BiTEs using the CD26 antibody sequence derived from YS110. The results of Examples 17 to 22 show that BiTEs of the present invention significantly suppress tumor growth in multiple animal tumor models, completely inhibiting tumor growth in some models, demonstrating significantly greater efficacy than BiTEs using the CD26 antibody sequence derived from YS110.

[0029] Furthermore, compared to BiTEs derived from other murine antibodies, BiTEs engineered based on the variable region of murine antibody 62 of the present invention exhibit greater safety and a lower risk of inducing a cytokine storm (Example 13).

[0030] Second, the bispecific T cell-inducing antibodies of the present invention that target CD26 are more effective and safer than existing anti-CD26 full-length monoclonal antibodies.

[0031] Regarding efficacy, the data in Example 11 showed that the bispecific antibody having a BiTE structure of the present invention has a superior killing effect against multiple CD26-positive tumor cells compared to the full-length IgG antibody YS110. The data in Examples 23 and 24 showed that the bispecific antibody having a BiTE structure of the present invention has a significantly superior tumor growth inhibitory effect compared to the full-length IgG antibody YS110 in multiple animal tumor models.

[0032] Regarding safety, as shown in Example 12, a full-length IgG anti-CD26 antibody cannot effectively induce T cell activation, whereas the CD26-CD3 bispecific antibody of the present invention in a BiTE structure activates T cells surrounding the tumor tissue and kills tumor cells after binding to them, without apparently inducing T cell activation before reaching the tumor site. The data in Examples 23 and 24 demonstrate that the bispecific antibody of the present invention in a BiTE structure has significantly superior safety to the full-length IgG antibody YS110 in various animal tumor models.

[0033] Third, compared with existing first-line treatments for related tumors (e.g., treatment of kidney cancer with BAVENCIO), the bispecific T cell-engaging antibodies of the present invention targeting CD26 have demonstrated more pronounced efficacy and safety (Examples 15, 22, and 23). [Brief explanation of the drawings]

[0034] [Figure 1] FIG. 1 shows the activity of humanized bispecific antibodies measured by dot blot assay. (a) shows the activity of humanized bispecific antibodies 18G272 and 18G278 measured by dot blot assay, and (b) shows the activity of humanized bispecific antibody 19G294 measured by dot blot assay. [Figure 2]1 is a graph showing the results of measuring the binding specificity between an antibody and a target cell antigen by flow cytometry. [Figure 3] 1 is a graph showing the simultaneous binding of 19G294 to CD26 protein and CD3 protein. [Figure 4] 1 is a graph showing the cytotoxic effect of humanized bispecific antibodies on PBMC cells. [Figure 5] 1 is a graph showing the tumor volume of animals in the 18G272 administration group and the 19G294 administration group in a NOD / SCID mouse model in which PC-3 cells were subcutaneously implanted. [Figure 6] 1 is a graph showing the tumor weights of animals in the 18G272-administered group and the 19G294-administered group in a NOD / SCID mouse model in which PC-3 cells were subcutaneously implanted. [Figure 7] 1 is a graph showing the tumor volume of animals in the 18G272 administration group in a NOD / SCID mouse model in which NCI-H596 cells were subcutaneously implanted. [Figure 8] 1 is a graph showing the tumor volume of animals in a 19G294 administration group in a NOD / SCID mouse model in which NCI-H226 cells were subcutaneously implanted. [Figure 9] 1 is a graph showing the animal body weights of the 19G294 administration group in a NOD / SCID mouse model subcutaneously implanted with NCI-H226 cells. [Figure 10] 1 is a graph showing the tumor volume of animals in a 19G294 administration group in a NOD / SCID mouse model in which A498 cells were subcutaneously implanted. [Figure 11] 1 is a graph showing the tumor volume of animals in the 17G29 administration group and the 19G295 administration group in a NOD / SCID mouse model in which OS-RC-2 cells were subcutaneously implanted. [Figure 12] 1 is a graph showing the tumor volume of animals in the 18G272-administered group and the 19G295-administered group in a NOD / SCID mouse model in which OS-RC-2 cells were subcutaneously implanted. [Figure 13] 1 is a graph showing the animal survival rates in the 18G272-administered group and the 19G295-administered group in a NOD / SCID mouse model subcutaneously implanted with OS-RC-2 cells. [Figure 14]1 is a graph showing the tumor volume of animals in the 18G272 administration group in a NOD / SCID mouse model in which OS-RC-2 cells were subcutaneously implanted. [Figure 15] 1 is a graph showing the tumor volume of animals in a 19G294 administration group in a NOD / SCID mouse model in which OS-RC-2 cells were subcutaneously implanted. [Figure 16] 1 is a graph showing the tumor volume of animals in the 19G294-administered group and the 21G587-administered group in a NOD / SCID mouse model in which OS-RC-2 cells were subcutaneously implanted. [Figure 17] 1 is a graph showing the animal weights of the 19G294 administration group and the 21G587 administration group at different doses in a NOD / SCID mouse model subcutaneously implanted with OS-RC-2 cells. [Figure 18] 1 is a graph showing the tumor volume of animals in the 19G294 administration group, the BAVENCIO+Axitinib combination group, and other groups in a NOD / SCID mouse model subcutaneously implanted with OS-RC-2 cells. [Figure 19] 1 is a graph showing the tumor volumes of animals in the 19G294 administration group, the BAVENCIO+Axitinib combination group, and other groups on day 26 after inoculation in a NOD / SCID mouse model subcutaneously implanted with OS-RC-2 cells. [Figure 20] 1 is a graph showing the tumor volumes of animals in the 19G294 administration group, the BAVENCIO+Axitinib combination group, and other groups on day 26 after inoculation in a NOD / SCID mouse model subcutaneously implanted with A498 cells. [Figure 21] 1 is a graph showing tumor weights in animals in the 19G294 administration group, the BAVENCIO+Axitinib combination group, and other groups 65 days after inoculation in a NOD / SCID mouse model in which A498 cells were subcutaneously implanted. DETAILED DESCRIPTION OF THE INVENTION

[0035] Unless expressly defined herein, technical terms used herein have the meanings that are commonly understood by those skilled in the art. Singular words such as "a," "an," "the," etc. also include the plural of the respective referent.

[0036] The term "or" means "and / or" and may be used interchangeably with "and / or." The term "CD26" is also referred to as dipeptidyl peptidase 4 (DPP4). The amino acid sequence of human CD26 can be found in Genbank under accession number NP_001926.2, and its cDNA sequence can be found in Genbank under accession number NM_001935.3.

[0037] The term "conservative amino acid substitution" refers to a substitution of a new amino acid for an original amino acid that does not significantly alter the chemical, physical, and / or functional properties (e.g., binding affinity to CD26) of an antibody or fragment thereof. Conservative amino acid substitutions are well known in the art, and include, for example, conservative substitutions of Ala to Gly, Ser, Arg to Lys, His, and Asn to Gln, His.

[0038] The term "affinity" refers to the strength of the interaction between an antibody and an antigen. The variable region of an antibody interacts with an antigen through non-covalent forces; the more interactions, the stronger the affinity.

[0039] The term "antibody" refers to a family of immunoglobulins capable of noncovalent, reversible, and specific binding to corresponding antigens. For example, naturally occurring IgG antibodies are tetramers, comprising at least two heavy chains and two light chains interlinked by disulfide bonds. Each heavy chain consists of a heavy chain variable region (VH) and a heavy chain constant region (CH), with the heavy chain constant region consisting of three domains: CH1, CH2, and CH3. Each light chain consists of a light chain variable region (VL) and a light chain constant region (CL), with the light chain constant region consisting of a single domain: CL. VH and VL can be further subdivided into highly variable complementarity-determining regions (CDRs, also known as hypervariable regions) and more conserved framework regions (FRs). Each VH or VL consists of three CDRs and four FRs, arranged from the amino terminus to the carboxy terminus in the order FR1, CDR1, FR2, CDR2, FR3, CDR3, and FR4. HCDR1, HCDR2, and HCDR3 are the three heavy chain complementarity determining regions, and LCDR1, LCDR2, and LCDR3 are the three light chain complementarity determining regions. The positions of CDRs and framework regions can be determined using various numbering systems well known in the art, such as Kabat, Chothia, and IMGT. In this application, the positions are determined using IMGT. The heavy chain variable region (VH) and light chain variable region (VL) are responsible for antigen recognition, and their complementarity determining regions (CDRs) in particular usually have specificity for different epitopes of the antigen. The constant region is mainly responsible for effector function.

[0040] The term "antigen-binding fragment" refers to an antigen-binding fragment of an antibody, i.e., a fragment of an antibody that retains the ability to specifically bind to an antigen, such as a fragment that retains one or more CDR regions, including, but not limited to, Fab fragments, Fv fragments, diabodies, linear antibodies, single-chain antibodies, nanobodies, multispecific antibodies, etc.

[0041] The term "monoclonal antibody" refers to a population of essentially homogeneous antibodies, i.e., the amino acid sequences of the antibody molecules within the population are the same, except for minor variations that may occur naturally. In contrast, polyclonal antibodies typically contain a variety of antibodies with variable regions (especially complementarity-determining regions) that differ in amino acid sequence. Monoclonal antibodies can be obtained by methods known to those skilled in the art. In this application, monoclonal antibodies are obtained by hybridoma technology. This method is one of many known methods for obtaining monoclonal antibodies.

[0042] The term "murine antibody" refers to an antibody that contains only rat or mouse immunoglobulin sequences.

[0043] The term "chimeric antibody" refers to an antibody in which the variable regions are derived from non-human (e.g., murine) antibody sequences and other portions are derived from human antibody sequences.

[0044] The term "humanized antibody" refers to an antibody in which the CDR regions are derived from non-human (e.g., murine) antibody sequences and other portions are derived from human antibody sequences.

[0045] The term "bispecific antibody" refers to an antibody that has two specific binding sites for the same or different antigens.

[0046] The term "BiTE" refers to a type of bispecific antibody, whose full name is "bispecific T cell engager." It is formed by linking two single-chain variable fragments (scFvs) with antigen specificity via a linker. One scFv targets a tumor-associated surface antigen, while the other scFv targets CD3 on the T cell surface. The single-chain variable fragment (scFv) is formed by linking the heavy and light chain variable regions of an antibody via a linker. A "linker" is an amino acid sequence used to connect different protein fragments, and its design and selection have been the subject of much research. In BiTEs, the linker connects one heavy chain variable region and one light chain variable region to form the scFv, and also connects the two scFvs in tandem to allow for free rotation. The linker is generally composed of a GGGGS repeat structure (S represents serine and G represents glycine). Glycine's small molecular weight and short side chain increase its flexibility, while serine's relatively strong hydrophilicity enhances the hydrophilicity of the peptide chain. Optimizing the number of repeats not only helps ensure the correct conformation of the scFv light and heavy chains, but also influences the distance between the two scFvs, thereby promoting optimal interaction between T cells and target cells at the immune synapse (Source: Zhou Chong et al., Research Progress in Bispecific Single-Chain Antibodies (BiTE) [J]. Biological Journal, 2018). Commonly used linkers for linking the heavy and light chain variable regions of scFvs include KESGSVSSSEQLAQFRSLD, EGKSSGSGSESKST, GSTSGGGSGGGSGGGGSS (US20180326032), and GSTSGSGKPGSGEGSTKG (Preclinical Development of Bivalent Chimeric Antigen Receptors Targeting Both CD19 and CD22). In a specific embodiment of the present application, a flexible linker consisting of three repeats of GGGGS is used to link one heavy chain and one light chain to form an scFv, and a flexible linker of GGGGS is used to connect two scFVs in series.A person skilled in the art can select and determine a linker from known linkers through routine experiments and a limited number of experiments based on the teachings of existing techniques, and can also optimize the number of GGGGS repeats in the two types of flexible linkers, or use flexible linkers other than GGGGS.

[0047] CD3 is a component of T cell signaling. When BiTE molecules bind to both T cells and tumor cells, T cells are activated, promoting CD8+ T cells to directly secrete perforin and granzymes, and CD4+ T cells to secrete cytokines, which further attract and activate tumor cytotoxic T cells, thereby damaging tumor cells. CD3 antibody sequences can be selected from those known in the art, such as OKT-3, L2K, TR66, UCHT1, SP34, IORT3, Catumaxomab, Blinatumomab, Solitomab, and WBP3311_2.306.4 (patent CN201880061333.1). While the CD3 antibody sequences of Solitomab and WBP3311_2.306.4 are used in the specific examples of this application, those skilled in the art are not limited to the specific CD3 antibody sequences used in the specific examples of this application and can select and determine CD3 antibodies from other known CD3 antibodies through routine experimentation and a limited number of experiments.

[0048] The term "chimeric BiTE bispecific antibody" as used herein refers to a murine CD26 scFv linked to a human CD3 scFv via a linker.

[0049] The terms "administration," "administering," "treating," and "treatment" refer to the contact of an exogenous agent, therapeutic agent, or diagnostic agent with a subject's tissues, organs, cells, or biological fluids. "Treating" means slowing or halting the progression of clinical symptoms of a disease, or alleviating or ameliorating at least one physical parameter, or preventing the progression of a disease. [Example]

[0050] Example 1. Generation and screening of anti-human CD26 hybridoma cell lines Step 1: Animal immunization Based on the CD26 cDNA sequence published in GenBank (GenBank accession number: NM_001935.3), an rCD26 expression vector was designed, and a His tag was introduced at the 5' end of the codon. The entire gene was synthesized and ligated into the pCHO1.0 plasmid, followed by expression and purification in CHO-S host cells. Following a standard immunization procedure, female BALB / c mice were immunized with the prepared recombinant CD26 target protein. For details on immunization, please refer to the "Experimental Guide for Antibody Preparation and Use." Serum titers from immunized mice were tracked using an indirect ELISA method, and the mouse with the highest serum titer was selected for fusion experiments with its spleen cells and myeloma cells.

[0051] Step 2: Cell fusion (1) Preparation of spleen cells Immunized mice were blooded by eye removal, sacrificed by cervical dislocation, and immersed in 75% (v / v) alcohol for 10 minutes. The spleens were then removed and placed on a sterile operating table. The cells were thoroughly disrupted and passed through the cell strainer. The cells were then centrifuged several times in sterile 1640 medium (Gibco). The cells were resuspended to prepare a single-cell suspension, counted, and stored for later use.

[0052] (2) Preparation of cultured cells A single 8-10 week old female BALB / c mouse was selected, eyeball enucleated to obtain negative serum, and then sacrificed by cervical dislocation. The mouse was then immersed in 75% (v / v) alcohol for 10 minutes, and the abdominal skin was aseptically peeled off to expose the peritoneum. Approximately 10 mL of 1640HT medium (purchased from SIGMA) was injected into the abdominal cavity with a syringe, and the abdomen was gently massaged and blown several times. The medium containing the macrophages was aspirated and poured into 20% 1640HAT medium, which was set aside for later use. Take a 2-3 week old female BALB / c mouse, kill it by cervical dislocation, immerse it in 75% (v / v) alcohol for 10 minutes, aseptically remove the thymus, place it in a cell strainer, crush it, and pass it through the cell strainer. Place the obtained thymocytes in 20% 1640HAT medium containing the above-mentioned macrophages and set aside for later use.

[0053] (3).Cell fusion The mouse myeloma cell line SP2 / 0 in the logarithmic growth phase was selected, collected, and counted. 8 The above SP2 / 0 cell line was used in approximately 2 × 10 7 The resulting mixture was added to a fusion tube and mixed. The tube was centrifuged at 1000 rpm for 10 minutes, the supernatant discarded, and the precipitate was dispersed by gently rubbing the tube in the palm of the hand. Within 60 seconds, 1 mL of preheated PEG1450 (polyethylene glycol 1450, purchased from SIGMA) was added, initially slowly and gradually more rapidly. 30 mL of 1640 HT medium was added to terminate the mixture. The mixture was centrifuged at 1000 rpm for 10 minutes, the supernatant discarded, the precipitate was dispersed by gently rubbing the tube, and the mixture was added to the 20% 1640 HAT medium obtained in step 2. After thoroughly mixing the HAT medium, 200 μL was added to a 96-well cell culture plate at 37°C in a 5% CO2 cell culture box. After one week, the 20% 1640HAT medium was replaced with 10% 1640HT medium, and the supernatant was collected and assayed three days later.

[0054] Step 3: Screening of anti-CD26 specific hybridoma cell lines (1) Assay plate preparation: Dilute recombinant CD26 target protein to 1 μg / ml in CB coating solution and coat a 96-well ELISA enzyme-labeled plate with 100 μl / well. Coat overnight at 2-8°C, wash once, and tap dry. Block with 2% BSA in PBST buffer (200 μl / well) for 2 hours at 37°C, tap dry, and set aside for later use.

[0055] (2) Screening for positive clones: Add 100 μl of test cell culture supernatant to the above measurement plate, react at 37°C for 30 minutes, wash, tap dry, add 100 μl of diluted HRP-labeled goat anti-mouse IgG to the plate, react at 37°C for 30 minutes, wash, tap dry, add 100 μl of TMB coloring solution to the plate, and let the plate develop for 15 minutes at 37°C in the dark. Add 50 μl of 2M H2SO4 to each well to terminate the reaction, and measure the OD 450 The data was read at OD. 450 The positive clone line was selected and subjected to cell cloning screening. After three or four rounds of cloning screening, if the positive rate of the monoclonal cell line was 100%, it was confirmed as a stable expression cell line and the cell line was immobilized.

[0056] Step 4: Activity assessment of CD26 murine monoclonal antibodies In the 786-0 cell reaction system, 786-0 cells were labeled with a green fluorescent dye, Calcein-AM, and the 786-0 cells were cultured at a cell concentration of 6 × 10 5 50 μl of CD26 antibody was inoculated into each well of a U-shaped 96-well cell culture plate at a final concentration of 100 ng / ml, with three parallel wells for each group. 50 μl of CD26 antibody was added to the reaction well of each sample well at a final concentration of 100 ng / ml. 50 μl of medium was added to the blank control well as a control. 50 μl of Triton X-100 was added to the positive control well at a final concentration of 1%. The wells were then cultured in a 37°C culture chamber for 30 minutes, and then incubated at a concentration of 6×10 at an E / T ratio of 10:1. 6 PBMC cells (peripheral blood mononuclear cells) were added at 37°C under carbon dioxide for 5 hours. After the reaction was completed, the supernatant was collected by centrifugation and placed in a new 96-well plate. The plate was then centrifuged again, and 80 μl of the supernatant was collected and placed in a black 96-well plate. The cell lysis rate was measured using a microplate reader with an excitation wavelength of 470 nm and an emission wavelength of 515 nm. The formula for calculating the cell lysis rate was: (V sample -V vehicle control ) / (V TritonX-100 -V vehicle control ) × 100%. (where Vsample is the average fluorescence signal reading of the drug-treated group measured at the excitation and emission wavelengths, and V vehicle control is the average fluorescence signal reading of the blank control measured at the excitation and emission wavelengths, and V TritonX-100 is the average fluorescence signal reading of the positive control measured at the excitation and emission wavelengths. [Table 1]

[0057] The above results show that when the antibody concentration was 100 ng / ml and the reaction time was 5 hours, the cytolysis rate of target cells 786-0 mediated by murine antibody No. 62 in PBMCs was 53.1%, which is much higher than that of other murine monoclonal antibodies such as No. 72, No. 160, No. 51, and No. 212, and is already equivalent to the humanized antibody YS110. 786-0 is a human tumor cell line. Theoretically, a successfully humanized antibody will have a much higher affinity for human-derived cells and a stronger lytic activity against the tumor cells than the parent murine antibody. The murine antibody No. 62 screened in this application has a cytotoxic effect against human tumor cells equivalent to that of the humanized antibody YS110, which has already completed a phase II clinical trial, demonstrating the superiority of the murine antibody No. 62. Those skilled in the art can reasonably predict that a humanized antibody derived from the murine antibody No. 62 as a parent antibody will have a more cytotoxic effect against human tumor cells than the murine antibody No. 62 itself or YS110. This was also demonstrated in Example 11.

[0058] Example 2. Sequencing and screening of hybridoma cell line variable regions Step 1. Extraction of total RNA from CD26 hybridoma cells Hybridoma cell lines were subcultured in T75 culture flasks. When the cells reached approximately 90% confluence, they were digested and collected by centrifugation. Total RNA was extracted from the monoclonal hybridoma cell lines using an RNA extraction kit (purchased from Roche). The first strand of cDNA was then reverse transcribed and amplified using a cDNA reverse transcription kit (purchased from Thermo). The reaction product was stored at -20°C. For long-term storage, store at -70°C.

[0059] Step 2. PCR amplification of heavy and light chain variable region genes First-strand cDNA from hybridoma cells was used as a template. A 50 μl reaction mixture was prepared by adding 1 μl of cDNA, 5 μl of 10x PCR buffer, 1 μl each of upstream and downstream primers (25 pmol), 1 μl of dNTPs, 1 μl of 25 mmol / L MgCl2, and 39 μl of H2O. The mixture was initially denatured at 95°C for 10 minutes, followed by 1 μl of Taq enzyme. The PCR amplification was performed at 94°C. The reaction conditions were 94°C, denaturation at 58°C for 1 minute, annealing at 1 minute, and extension at 72°C for 1.5 minutes, for a total of 30 cycles, followed by a 10-minute hold at 72°C. Five μl of the PCR product was electrophoresed on a 1.2% agarose gel.

[0060] Step 3. Cloning and sequencing of heavy and light chain variable region genes The heavy and light chain variable region genes were ligated into the pGM-T vector according to the instructions of the pGM-T Fast Ligation Kit (Beijing Tiangen Biochemical Technology Co., Ltd., VT207-02), and then transformed into E. coli Top10 competent cells, subjected to blue-white selection, and cultured at 37°C for 12–16 hours.

[0061] The resulting white colonies were inoculated into 1-5 ml of LB medium containing 100 μl of ampicillin and shaken at 37°C for 3-4 hours. PCR was then used to screen for clones with the correct inserted sequence, and positive clones were simultaneously screened and sequenced. The sequencing results were compared and analyzed using the IMGT gene bank, and the amino acid sequences of the antibody heavy chain, light chain variable region, and CDR region were obtained.

[0062] Murine antibody No. 62 comprises an HCDR1 set forth in SEQ ID NO: 1, an HCDR2 set forth in SEQ ID NO: 2, an HCDR3 set forth in SEQ ID NO: 3, an LCDR1 set forth in SEQ ID NO: 4, an LCDR2 composed of Arg Met Ser (where Arg represents arginine, Met represents methionine, and Ser represents serine), and an LCDR3 set forth in SEQ ID NO: 5. Its heavy chain variable region is set forth in SEQ ID NO: 31, and its light chain variable region is set forth in SEQ ID NO: 32.

[0063] Example 3. Design of chimeric BiTE bispecific antibodies The murine monoclonal antibodies Nos. 72, 160, 62, and 51 obtained in Example 1 were sequenced by the method described in Example 2, and based on the obtained heavy and light chain sequences, bispecific antibodies targeting both CD26 and CD3 were designed.

[0064] The light chain variable region targeting the first antigen, CD26, was linked to the heavy chain variable region targeting CD26 by a short linking peptide (shown in SEQ ID NO: 12) to form a CD26 single-chain antibody domain: single-chain antibody derived from murine antibody #72 (shown in SEQ ID NO: 16), single-chain antibody derived from murine antibody #160 (shown in SEQ ID NO: 17), single-chain antibody derived from murine antibody #62 (shown in SEQ ID NO: 18), and single-chain antibody derived from murine antibody #51 (shown in SEQ ID NO: 19).

[0065] The sequence of the single-chain antibody targeting the second antigen, CD3 protein, was formed by linking the heavy chain variable region (shown in SEQ ID NO: 10) and the light chain variable region (shown in SEQ ID NO: 11) with a short linking peptide (shown in SEQ ID NO: 12).

[0066] Four types of first recognition domain sequences and one type of second recognition domain sequence were fused with a connecting short peptide (shown in SEQ ID NO: 13) to form a bispecific antibody with a BiTE structure, resulting in four BiTE sequences: 17G105 (shown in SEQ ID NO: 20) derived from murine antibody #72, 17G108 (shown in SEQ ID NO: 21) derived from murine antibody #160, 17G61 (shown in SEQ ID NO: 22) derived from murine antibody #62, and 17G131 (shown in SEQ ID NO: 23) derived from murine antibody #51.

[0067] Example 4. Construction, expression and purification of expression plasmids for anti-CD26-CD3 chimeric bispecific antibodies The amino acid sequence of the CD26 single-chain antibody (SEQ ID NO: 16-19) was optimized based on the codon preference of the mammalian CHO cell line to obtain the optimized gene sequence of the CD26 single-chain antibody (SEQ ID NO: 25-28), with an AvrII restriction enzyme site and a Kozak sequence inserted upstream. The amino acid sequence of the solitomab CD3 single-chain antibody was codon-optimized to obtain the optimized gene sequence (SEQ ID NO: 29), with a linking short peptide gene (SEQ ID NO: 30) inserted upstream and a stop codon and BstZ17I restriction enzyme site inserted downstream. The optimized AvrII restriction enzyme site, Kozak sequence, signal peptide, and CD26 single-chain antibody sequence were fused to the CD3 single-chain antibody, stop codon, and BstZ17I restriction enzyme site via a linking peptide to form four chimeric BiTE genes. These genes were then synthesized and constructed into the pUC57 plasmid to form several long-term storage plasmids, named pUC57-17G105, pUC57-17G108, pUC57-17G61, and pUC57-17G131, respectively.

[0068] The target gene was amplified and the PCR product was recovered by 1% agarose gel electrophoresis. The final target gene was double-digested with AvrII and BstZ17I. The PCR product and the pZHK2.0 vector were then digested with AvrII and BstZ17I. The vector backbone pCHO 1.0 was purchased from Invitrogen, and "expression cassette 1" was digested with the restriction enzyme SfiI and then cyclized with T4 ligase to form the pZHK2.0 vector. The PCR product was double-digested with T4 ligase and ligated into the pZHK 2.0 vector, then transformed into Top10 competent cells, plated on a kanamycin-tolerant LB plate, and grown overnight at 37°C. The next day, positive clones were screened and sequenced for comparison; their sequences were found to be completely consistent with the predicted sequence.

[0069] A mammalian cell line stably expressing a CD26-CD3 chimeric BiTE bispecific antibody was inoculated into Dynamis medium (A2617501, purchased from Thermo Fisher Scientific) and cultured in a fed-batch format at 37°C, 8% CO2, and 130 rpm. The culture supernatant was collected and centrifuged at 12,000 rpm for 15 minutes at low temperature. The supernatant was then filtered through a 0.45 μm collection filter membrane to obtain the treated culture supernatant, which was then purified by chromatography. The size of the target product obtained was approximately 55 kDa.

[0070] Example 5. Measurement of affinity of anti-CD26-CD3 chimeric bispecific antibody to CD26 protein and CD3 protein Fortebio Intermolecular Interaction Octet QK e The affinity of the chimeric antibody to the antigen protein CD26 was measured using this system. First, the amino-coupling biosensor AR2G was activated with EDC / s-NHS. The activated biosensor AR2G was immobilized in a CD26 protein solution diluted with 10 mM sodium acetate (pH 5.0). The CD26 protein-immobilized biosensor was quenched in 1 M ethanolamine (pH 8.5) and equilibrated in 1x kinetic solution. After equilibration, the biosensor was bound to a humanized bispecific antibody solution and then dissociated in 1x kinetic solution. Data analysis was performed using Fortebio Data Analysis 8.0 software to calculate the affinity constant (KD), association rate constant (k), and dissociation rate constant (kdis). The response value indicates the number of antibody molecules capable of binding to the antigen, k indicates the rate at which the antibody binds to the corresponding receptor, and k indicates the rate at which the antibody dissociates from the corresponding receptor. NA indicates that no data was available. [Table 2]

[0071] The above data demonstrate that chimeric antibodies can bind and dissociate from CD26 protein. 17G61 (anti-CD26-CD3 chimeric bispecific antibody derived from murine antibody #62) had the highest binding response to CD26 protein, indicating that more 17G61 molecules bound to CD26 protein at the same concentration. 17G105 (anti-CD26-CD3 chimeric bispecific antibody derived from murine antibody #72) and 17G131 (anti-CD26-CD3 chimeric bispecific antibody derived from murine antibody #51) also bound to CD26 protein, but their binding responses were significantly lower than those of 17G61. The smaller KD value between 17G61 and CD26 indicates a stronger affinity between 17G61 and CD26 protein.

[0072] Fortebio Intermolecular Interaction Octet QK e The affinity of the chimeric antibody to the antigen protein CD3 was measured using this system. First, the amino-coupling biosensor AR2G was activated with EDC / s-NHS. The activated biosensor AR2G was immobilized in a CD3 protein solution diluted with 10 mM sodium acetate (pH 5.0). The biosensor with immobilized CD3 protein was quenched in 1 M ethanolamine (pH 8.5) and equilibrated in 1x kinetic solution. After equilibration, the antibody was bound to a humanized bispecific antibody solution and then dissociated in 1x kinetic solution. Data analysis was performed using Fortebio Data Analysis 8.0 software, and affinity constants (KD) were calculated. [Table 3]

[0073] From the above data, it can be seen that 17G61 and CD3 protein have the highest binding response. This indicates that more 17G61 molecules bind to CD3 protein under the same concentration conditions. 17G105 binds to CD3 protein, but its binding response is lower than that of 17G61. The above research results indicate that 17G61 binds and dissociates well with CD3 protein.

[0074] Example 6. Humanization of anti-CD26 murine antibodies Based on the sequences of candidate murine antibodies No. 62 and No. 212, which are shown in Example 1 to have high binding affinity to CD26 and high killing activity against CD26-highly expressing tumor cells, humanization design was performed. Humanization analysis was performed on the variable regions of the heavy chain (shown in SEQ ID NO: 31) and light chain (shown in SEQ ID NO: 32) of murine antibody No. 62, and the variable regions of the heavy chain (shown in SEQ ID NO: 33) and light chain (shown in SEQ ID NO: 34) of murine antibody No. 212. Through calculation and prediction, multiple humanized heavy and light chain pairs were obtained from each antibody. Murine antibody 62 was humanized to obtain four different pairs of heavy and light chain amino acid sequences: 18G272-VL (SEQ ID NO:7), 18G272-VH (SEQ ID NO:6), 19G292-VL (SEQ ID NO:35), 19G292-VH (SEQ ID NO:36), 19G293-VL (SEQ ID NO:37), 19G293-VH (SEQ ID NO:38), 19G294-VL (SEQ ID NO:9), and 19G294-VH (SEQ ID NO:8).

[0075] The murine antibody 212 was humanized to obtain four different pairs of heavy and light chain amino acid sequences: 18G278-VL (SEQ ID NO: 39), 18G278-VH (SEQ ID NO: 40), 19G295-VL (SEQ ID NO: 41), 19G295-VH (SEQ ID NO: 42), 19G296-VL (SEQ ID NO: 43), 19G296-VH (SEQ ID NO: 44), 19G297-VL (SEQ ID NO: 45), and 19G297-VH (SEQ ID NO: 46). The humanized antibody YS110 was used as a positive control, and the heavy and light chain numbers and sequences thereof are as follows: YS110-VL (SEQ ID NO: 47), YS110-VH (SEQ ID NO: 48).

[0076] Example 7. Expression plasmid construction, stable expression and purification of humanized bispecific antibodies The light and heavy chain pairs of the above humanized CD26 antibodies (18G272, 19G292, 19G293, 19G294, 18G278, 19G295, 19G296, and 19G297) were each linked by a short linking peptide (SEQ ID NO: 12), and a short linking peptide (SEQ ID NO: 13) was added downstream. The sequences were then optimized based on the CHO codon preference of mammalian cells to obtain the optimized CD26 single-chain antibody gene sequences (SEQ ID NOs: 49-56). An AvrII restriction enzyme site and a Kozak sequence were added upstream, and the CD3 single-chain antibody gene (SEQ ID NO: 29), a stop codon, and a BstZ17I restriction enzyme site were added downstream. The humanized CD26-CD3 BiTE gene sequences were obtained, directly synthesized, and constructed into pUC57 plasmids, named pUC57-18G272, pUC57-19G292, pUC57-19G293, pUC57-19G294, pUC57-18G278, pUC57-19G295, pUC57-19G296, and pUC57-19G297.

[0077] At the same time, the light and heavy chains of the YS110 antibody were linked by a short linking peptide (SEQ ID NO: 12), followed by a short linking peptide (SEQ ID NO: 13) downstream, an AvrII restriction enzyme site and a Kozak sequence upstream, and the CD3 single-chain antibody gene (SEQ ID NO: 29), a stop codon, and a BstZ17I restriction enzyme site downstream. The humanized CD26-CD3 BiTE gene sequence was obtained, directly synthesized, and constructed into the PUC57 plasmid, designated pUC-17G29.

[0078] The light and heavy chains of 19G294 were linked by a short linker peptide (SEQ ID NO: 12), and a short linker peptide (SEQ ID NO: 13) was added downstream. This was then linked to a single-chain antibody (SEQ ID NO: 57) derived from the CD3 antibody WBP3311_2.306.4, which was then linked by a short linker peptide (SEQ ID NO: 12). This single-chain antibody was optimized for CHO codon preference, and an AvrII restriction enzyme site and a Kozak sequence were introduced upstream, and a stop codon and a BstZ17I restriction enzyme site were added downstream. This single-chain antibody was then synthesized directly and constructed into the pUC57 plasmid, designated pUC57-21G587.

[0079] The target gene was amplified and the PCR product was recovered by 1% agarose gel electrophoresis. The PCR product and the pZHK2.0 vector were double-digested with AvrII and BstZ17I, and the double-digested PCR product was ligated with T4 ligase into the pZHK2.0 vector. The cells were then transformed into Top10 competent cells, plated onto kanamycin-tolerant LB plates, and grown overnight at 37°C. The next day, positive clones were screened and sequenced. The sequences were found to be identical to the predicted sequence. This indicates that the humanized CD26-CD3 BiTE bispecific antibody expression plasmid was obtained.

[0080] A mammalian cell line stably expressing the CD26-CD3 humanized BiTE bispecific antibody was inoculated into Dynamis medium and cultured in a fed-batch format at 37°C, 8% CO2, and 130 rpm. The culture supernatant was collected and centrifuged at 12,000 rpm for 15 minutes at low temperature. The supernatant was then filtered through a 0.45 μm collection filter membrane to obtain the treated culture supernatant, which was then purified by chromatography. The size of the target product obtained was approximately 55 kDa, which was accurate.

[0081] Example 8. Dot blot analysis of purified humanized bispecific antibody samples The purified samples were diluted to different concentrations, and 2 μL of each diluted sample was spotted onto the NC membrane. The positive control sample was 17G29. Once fully absorbed, the membrane was blocked with 5% nonfat dry milk blocking solution at room temperature on a horizontal shaker for 1 hour, followed by three 5-minute washes with TBST. The CD3-HRP protein was added to 2.5% nonfat dry milk blocking solution at a 1:1000 ratio (50 μL of CD3-HRP was added to 50 mL of the prepared 2.5% nonfat dry milk blocking solution). The membrane was shaken at room temperature on a horizontal shaker for 1 hour, followed by three 5-minute washes with TBST. The CD26-HRP protein was added to 2.5% nonfat dry milk blocking solution at a 1:1000 ratio (50 μL of CD26-HRP was added to 50 mL of the prepared 2.5% nonfat dry milk blocking solution). The membrane was shaken at room temperature on a horizontal shaker for 1 hour, followed by three 5-minute washes with TBST. The TBST solution on the membrane was blotted dry with a paper towel, and then a coloring solution (SuperSignal West Pico Chemiluminescent Substrate) was added. The coloring solution was developed for 1 minute, blotted dry with a paper towel, and the membrane was placed in a gel imaging system for exposure. The continuous exposure program was selected, and the settings were 15 seconds per frame, with 12 consecutive frames. Dot blot analysis of 18G272 (humanized murine antibody #62) and 18G278 (humanized murine antibody #212) showed that the binding activity of 18G272 to CD26 and CD3 at different concentrations was similar to that of the positive control 17G29 (an anti-CD26-CD3 humanized bispecific antibody derived from YS110), but significantly higher than that of 18G278 (Figure 1(a)). A total of six samples, 19G292-19G297 (anti-CD26-CD3 humanized bispecific antibody derived from murine antibody no. 62), were assayed by dot blot. The results showed that all 19G292-19G297 had binding activity to CD26 or CD3, and the binding activity of 19G292-19G296 to CD26 was close to that of the positive control 17G29, with 19G294 in particular being even better than the positive control (Figure 1(b)).

[0082] Example 9: Study of the binding specificity of humanized bispecific antibodies to target cell antigens 1. Screening for CD26-positive and CD26-negative cells Target cells were cultured in T75 cell culture flasks, and when the cells were more than 80% confluent, they were harvested by trypsin digestion, washed once with PBS, and counted using a hemocytometer to obtain 5 × 10 cells. 5 The cells were divided into portions. Anti-CD26 monoclonal antibody was used as the primary antibody and incubated with target cells at room temperature for approximately 40 minutes. After incubation, the cells were centrifuged to discard the supernatant, the cell pellet was resuspended in PBS, centrifuged again to discard the supernatant, and the cell pellet was collected. The cell pellet was then resuspended with the corresponding Alexa Fluor 488 murine anti-human IgG1 antibody as the secondary antibody and incubated at room temperature, protected from light, for approximately 30 minutes. After incubation, the cells were washed twice with PBS, centrifuged to discard the supernatant, and the cell pellet was collected. The cell pellet was resuspended in approximately 200 μl of PBS solution and analyzed by flow cytometry to measure CD26 positivity within 1 hour.

[0083] Flow cytometry revealed CD26-positive cell lines, including 786-0, OS-RC-2, A498, NCI-H226, NCI-H2052, NCI-H596, HCC827, Huh-7, and PC-3. CD26-negative cell lines included G401 and A-375. The positivity rates are shown in Table 4. The CD26-positive cell lines were consistent with those described in the reference literature (Chinese Patent Application No. CN200680034937.4, CD26 / DPP4-a potential biomarker and target for cancer therapy, Pharmacology & Therapeutics (2019), 198:135-159). [Table 4]

[0084] 2. FITC-labeled antibody: Fluorescein (FITC) was weighed out and dissolved in DMSO to a final concentration of 1 mg / ml. 500 μl of 18G272 antibody was taken separately, adjusted to a final concentration of 1 mg / ml, and 1 / 10 the volume of 1M Na2CO3 was added and mixed thoroughly. 15 μl of FITC solution was then added, mixed thoroughly, and incubated at room temperature in the dark for 3.5 hours. After incubation, the mixture was centrifuged in a 10 kDa ultrafiltration tube to thoroughly remove FITC that was not bound to 18G272. The solution in the tube was then collected.

[0085] 3. Measurement using a flow cytometer Cells cultured in a T75 cell culture flask to about 80% capacity were digested with trypsin, collected, resuspended in an appropriate amount of PBS, and counted using a hemocytometer. 5 The cells were divided into portions, centrifuged to collect the cell pellet, resuspended in FITC-labeled 18G272, placed in a homogenizer, and incubated at room temperature for 30 minutes in the dark. After incubation, the cells were resuspended in PBS, centrifuged, and the supernatant discarded. The same procedure was repeated once, and the cell pellet was resuspended in 500 μl of PBS solution. Negative control: The other sample was treated in parallel with the above procedure without adding the antibody sample. The resulting solution was used as a negative control for FITC labeling. The positive rate of FITC labeling was analyzed using a BD Accuri C6 flow cytometer.

[0086] As a result, the humanized bispecific antibody 18G272 bound to CD26-positive cells such as 786-0 cells, NCI-H226 cells, PC-3 cells, and HCC827 cells, but did not bind to CD26-negative A375 cells. The binding of 18G272 to the CD26 target is specific.

[0087] Using the same method, the binding specificity of humanized bispecific antibody 19G294 to target cell antigens was measured. 19G294 bound to CD26-positive cells, such as 786-0 cells, NCI-H226 cells, PC-3 cells, and OS-RC-2 cells, but not to CD26-negative G401 cells. Humanized bispecific antibody 19G294 bound to CD26-positive cells but not to CD26-negative cells, demonstrating its specific binding to the CD26 antigen target. Representative graphs showing the binding specificity of antibodies to target cell antigens measured by flow cytometry are shown in FIG. 2, and detailed data are shown in Tables 5 and 6. [Table 5] [Table 6] Other bispecific antibodies of the invention also bind to CD26-positive cells and not to CD26-negative cells, and are specific in their binding to their CD26 antigen target.

[0088] Example 10: Affinity analysis of humanized bispecific antibodies with CD26 protein and CD3 protein 1. Study of affinity with CD26 protein First, CD26 protein was biotinylated with biotin EZ-link NHS-PEG12-Biotin. The biotinylated CD26 protein was immobilized on an SA biosensor, equilibrated in 1x Kinetics solution, and then bound to the antibody molecule solution to be measured. The antibody concentration gradients were 31.3 nM, 62.5 nM, 125 nM, 250 nM, and 500 nM. The antibody was then dissociated in 1x Kinetics solution. Data were analyzed using Fortebio Data Analysis 8.0 software, and affinity constants were calculated. [Table 7]

[0089] The above data show that, with the same anti-CD3 moiety, the affinity constants of the humanized antibodies derived from murine antibody 62 (19G292, 19G293, and 19G294) are all lower than those of the humanized antibodies derived from murine antibody 212 or YS110, but they have higher affinity for CD26 than the former. On the other hand, 19G294 and 21G587 have the same anti-CD26 moiety but different anti-CD3 moieties, but both have similar levels of affinity for CD26 protein.

[0090] 2. Study of the affinity of antibodies to CD3 protein First, CD3 protein was biotinylated with biotin EZ-link NHS-PEG12-Biotin. The biotinylated CD3 protein was immobilized on an SA biosensor, equilibrated in 1x Kinetics solution, and then bound to a humanized bispecific antibody solution. The measured concentration gradient of the humanized bispecific antibody was 200 nM, 400 nM, 800 nM, and 1600 nM. The antibody was then placed in 1x Kinetics solution and dissociated. Data were analyzed using Fortebio Data Analysis 8.0 software, and affinity constants were calculated. [Table 8]

[0091] In this experiment, the affinity constant of 19G294 for CD3 protein was 0.751 nM. 19G296 and 19G297 bound to CD3 with relatively fast dissociation rates. Humanized bispecific antibodies 19G294, 19G292, and 19G293, derived from murine antibody 62, which have the same anti-CD3 moiety, exhibited significantly higher affinity for CD3 protein than humanized bispecific antibody 17G29, derived from YS110. Meanwhile, 19G294 and 21G587, which have the same anti-CD26 moiety but different anti-CD3 moieties, exhibited comparable levels of affinity for CD3 protein.

[0092] The affinity data for the above antibodies with CD26 protein (Table 7) and CD3 protein (Table 8) show that the humanized antibodies derived from murine antibody #62 (19G292, 19G293, 19G294) have better CD26 protein binding avidity than the humanized antibodies derived from murine antibody #212 (19G296, 19G297). The humanized antibodies derived from murine antibody #62 (19G292, 19G293, 19G294) exhibit better CD3 protein binding and dissociation avidity than the humanized antibodies derived from murine antibody #212 (19G296, 19G297) and the humanized antibody YS110 (17G29), making 19G294 a more advantageous combination of bispecific antibody molecules.

[0093] 3. Observation of simultaneous binding of bispecific antibodies to CD26 and CD3 proteins First, 19G294 protein was biotinylated with Biotin EZ-link NHS-PEG12-Biotin. The biotinylated 19G294 protein was immobilized on an SA biosensor, equilibrated in 1x Kinetics solution, and then bound to either a 400 nM CD26 protein solution and then a 400 nM CD3 solution containing CD26 protein (400 nM), or a 400 nM CD3 protein solution and then a 400 nM CD3 solution containing CD26 protein (400 nM). Binding status was monitored using Fortebio Data Analysis 8.0 software.

[0094] As shown in Figure 3, 19G294 immobilized on the SA sensor can continue to bind to CD3 protein in a CD26 solution even after its binding to CD26 has reached saturation (Fig. 3, Sensor G9 curve). Also, even after its binding to CD3 has reached saturation, it can continue to bind to CD26 protein in a CD3 solution (Fig. 3, Sensor H9 curve). Figure 3, Sensor F9, is a negative control, demonstrating the absence of nonspecific binding between the sensor and the solution. These results demonstrate that 19G294 can continue to bind to either CD26 or CD3 protein even after first binding to either CD26 or CD3, indicating that there is no order in binding.

[0095] Example 11 Evaluation of humanized bispecific antibody-mediated cytotoxicity of PBMCs to target cells 1. Human renal carcinoma cell line 786-0 model In the 786-0 cell reaction system, 786-0 cells were labeled with a green fluorescent dye, Calcein-AM, at a cell concentration of 6 × 10 5 50 μl of 786-0 cells were seeded per well into a U-shaped 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of humanized bispecific antibodies were added to the corresponding sample reaction wells at final concentrations of 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of medium was added to the blank control wells. 50 μl of Triton X-100 (1% Triton X-100) was added to the positive control wells at a final concentration of 1%. Furthermore, 6 × 10 mAb was added to the positive control wells at an E / T ratio of 10:1. 6 50 μl of PBMC cells (cells / ml) was added, and the reaction system was incubated at 37°C under carbon dioxide for 5 hours. After the reaction was completed, the supernatant was collected by centrifugation and placed in a new 96-well plate. The plate was then centrifuged again, and 80 μl of the supernatant was collected and placed in a black 96-well plate. The results were measured using a microplate reader under conditions of an excitation wavelength of 470 nm and an emission wavelength of 515 nm.

[0096] 2. Human renal carcinoma cell line OS-RC-2 model In the OS-RC-2 cell reaction system, OS-RC-2 cells were labeled with a green fluorescent dye, Calcein-AM, at a cell concentration of 6 × 10 5 50 μl of OS-RC-2 cells were seeded per well into a U-shaped 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of humanized bispecific antibodies were added to the corresponding sample reaction wells at final concentrations of 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of medium was added to the blank control wells. 50 μl of Triton X-100 (1% Triton X-100) was added to the positive control wells at a final concentration of 1%. Furthermore, 50 μl of Triton X-100 (9×10 Triton X-100) was added to the positive control wells at an E / T ratio of 15:1. 6 50 μl of PBMC cells (cells / ml) was added, and the reaction system was incubated at 37°C under carbon dioxide for 5 hours. After the reaction was completed, the supernatant was collected by centrifugation and placed in a new 96-well plate. The plate was then centrifuged again, and 80 μl of the supernatant was collected and placed in a black 96-well plate. The results were measured using a microplate reader under conditions of an excitation wavelength of 470 nm and an emission wavelength of 515 nm.

[0097] 3. Human mesothelioma NCI-H226 cell model In the NCI-H226 cell reaction system, NCI-H226 cells were labeled with a green fluorescent dye, Calcein-AM, at a cell concentration of 6 × 10 5 50 μl of NCI-H226 cells were seeded per well into a U-shaped 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml of humanized bispecific antibody was added to the corresponding sample reaction wells at final concentrations of 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml, respectively. 50 μl of medium was added to the blank control wells. 50 μl of Triton X-100 was added to the positive control wells at a final concentration of 1% as a positive control. Furthermore, 9 × 10 mAb was added to the positive control wells at an E / T ratio of 15:1. 650 μl of PBMC cells (cells / ml) was added, and the reaction system was incubated at 37°C under carbon dioxide for 5 hours. After the reaction was completed, the supernatant was collected by centrifugation and placed in a new 96-well plate. The plate was then centrifuged again, and 80 μl of the supernatant was collected and placed in a black 96-well plate. The results were measured using a microplate reader under conditions of an excitation wavelength of 470 nm and an emission wavelength of 515 nm.

[0098] 4. Human mesothelioma NCI-H2052 cell model In the NCI-H2052 cell reaction system, NCI-H2052 cells were labeled with a green fluorescent dye, Calcein-AM, at a cell concentration of 6 × 10 5 50 μl of NCI-H2052 cells were seeded per well into a U-shaped 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml of humanized bispecific antibody was added to the corresponding sample reaction wells at final concentrations of 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml, respectively. 50 μl of medium was added to the blank control wells. 50 μl of Triton X-100 was added to the positive control wells at a final concentration of 1% as a positive control. Furthermore, 50 μl of Triton X-100 was added to the positive control wells at a concentration of 9 × 10 at an E / T ratio of 15:1. 6 50 μl of PBMC cells (cells / ml) was added, and the reaction system was incubated at 37°C under carbon dioxide for 5 hours. After the reaction was completed, the supernatant was collected by centrifugation and placed in a new 96-well plate. The plate was then centrifuged again, and 80 μl of the supernatant was collected and placed in a black 96-well plate. The results were measured using a microplate reader under conditions of an excitation wavelength of 470 nm and an emission wavelength of 515 nm.

[0099] 5. Human prostate cancer PC-3 cell model In the PC-3 cell reaction system, PC-3 cells were labeled with a green fluorescent dye, Calcein-AM, at a cell concentration of 6 × 10 550 μl of PC-3 cells were seeded per well into a U-shaped 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of humanized bispecific antibodies were added to the corresponding sample reaction wells at final concentrations of 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of medium was added to the blank control wells. 50 μl of Triton X-100 (1% Triton X-100) was added to the positive control wells at a final concentration of 1%. Furthermore, 50 μl of 9×10 mAb was added to the positive control wells at an E / T ratio of 15:1. 6 50 μl of PBMC cells (cells / ml) was added, and the reaction system was incubated at 37°C under carbon dioxide for 5 hours. After the reaction was completed, the supernatant was collected by centrifugation and placed in a new 96-well plate. The plate was then centrifuged again, and 80 μl of the supernatant was collected and placed in a black 96-well plate. The results were measured using a microplate reader under conditions of excitation wavelength of 470 nm and emission wavelength of 515 nm.

[0100] 6. Human non-small cell lung cancer (HCC) 827 cell model In the HCC 827 cell reaction system, HCC 827 cells were labeled with a green fluorescent dye, Calcein-AM, at a cell concentration of 6 × 10 5 50 μl of HCC 827 cells were seeded per well into a U-shaped 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of humanized bispecific antibodies were added to the corresponding sample reaction wells at final concentrations of 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml, respectively. 50 μl of medium was added to the blank control wells. 50 μl of Triton X-100 (1% Triton X-100) was added to the positive control wells at a final concentration of 1%. Furthermore, 6 × 10 mAb was added to the positive control wells at an E / T ratio of 10:1. 6 50 μl of PBMC cells (cells / ml) was added, and the reaction system was incubated at 37°C under carbon dioxide for 5 hours. After the reaction was completed, the supernatant was collected by centrifugation and placed in a new 96-well plate. The plate was then centrifuged again, and 80 μl of the supernatant was collected and placed in a black 96-well plate. The results were measured using a microplate reader under conditions of an excitation wavelength of 470 nm and an emission wavelength of 515 nm.

[0101] 7. Experimental results Cell lysis rate=(V sample -V vehicle control ) / (V TritonX-100 -V vehicle control ) × 100%. (where V sample is the mean value of the fluorescent signal readings of the drug-treated group, and V vehicle control is the mean fluorescence signal reading of the solvent control group, and V TritonX-100 is the average fluorescence signal reading of the positive control group.) The cytolysis rate and sample concentration values ​​were calculated using the software GraphPad Prism 7.00 to determine the IC of the sample-mediated PBMC target cells. 50 The value was calculated. [Table 9]

[0102] Anti-CD26-CD3 humanized bispecific antibodies derived from murine antibody #62, such as 18G272, 19G292, 19G293, and 19G294, were able to induce cytotoxic effects against multiple types of CD26-positive tumor cells by PBMCs and had killing activity in vitro. Anti-CD26-CD3 humanized bispecific antibodies derived from murine antibody #212, such as 19G295, 19G296, and 19G297, were unable to induce cytotoxic effects against CD26-positive tumor cells by PBMCs.

[0103] These results demonstrate that the murine antibodies screened in this study have superior killing activity against CD26-positive tumor cells, and that the anti-CD26-CD3 humanized antibody derived from murine antibody No. 62 has superior killing activity against multiple types of CD26-positive tumors compared to 17G29 (anti-CD26-CD3 humanized antibody derived from murine antibody No. 62) and anti-CD26-CD3 humanized antibody derived from murine antibody No. 212. Furthermore, it demonstrates that the bispecific antibodies with BiTE structure of this study (18G272, 19G292, 19G293, 19G294, and 17G29) have superior killing activity against multiple types of CD26-positive tumors compared to the full-length IgG antibody YS110.

[0104] Example 12 Activation of PBMC cells during the cytotoxic effect of PBMC on target cells via a humanized bispecific antibody PBMC cells + bispecific antibody system: PBMC cells were collected and counted, and 1 × 10 6 PBMC cells + bispecific antibody + OS-RC-2 cell line: PBMC cells were collected and counted, and 6 x 10 5 The antibodies were resuspended at a final concentration of 1 ng / ml and added to a 12-well cell culture plate at 100 ul / well. The antibodies were diluted to a final concentration of 1 ng / ml and added to 9 x 10 6 cells / ml and added at 100 ul / well to corresponding 12-well cell culture plates and co-incubated for 24 hours.

[0105] After incubation, the cells were centrifuged to collect the cell pellet, which was then resuspended in PBS. Anti-CD25-APC antibody (purchased from Miltenyl) and anti-CD69-PE antibody were added, and the cells were incubated in a refrigerator at 4°C for 10 minutes. After incubation, the cells were washed twice with PBS, resuspended in PBS, and then analyzed using a BD ACCURI C6 flow cytometer.

[0106] In the "PBMC cells + bispecific antibody" system, the bispecific antibody did not induce increased expression of CD25 and CD69 on PBMC cells. In the "PBMC cells + bispecific antibody + OS-RC-2 cells" system, the bispecific antibody induced increased expression of CD25 and CD69 on PBMC cells and promoted PBMC cell activation. In the "PBMC cells + YS110 + OS-RC-2 cells" system, YS110 failed to induce increased expression of CD25 and CD69 on PBMC cells. The results are shown in Table 10. [Table 10]

[0107] A similar phenomenon was observed with other CD26-positive tumor cells such as 786-0. That is, in the "PBMC cells + bispecific antibody" system, the bispecific antibody did not induce increased expression of CD25 and CD69 in PBMC cells, but in the "PBMC cells + bispecific antibody + other CD26-positive tumor cells (e.g., 786-0)" system, it induced increased expression of CD25 and CD69 in PBMC cells.

[0108] CD25 and CD69 are important markers for T cell activation. These results demonstrate that the full-length IgG antibody YS110 cannot effectively activate T cells, and that the CD26-CD3 bispecific antibody does not significantly activate T cells before reaching the tumor site. It is only after binding to tumor cells that the antibody activates T cells surrounding the tumor tissue and subsequently kills tumor cells, improving the safety of its action.

[0109] Example 13 Measurement of cytokines during the course of humanized bispecific antibody-mediated cytotoxic effect on target cells by PBMC 6 × 10 OS-RC-2 cells 5 PBMC cells from a single donor were collected and resuspended at 9 x 10 cells / ml and added to a 12-well plate at 100 μl per well. 6 The cells were resuspended at 100 μl per well, and 100 μl of 19G294, 18G272, and 17G29 samples at 0 ng / ml and 3 ng / ml were added to each well for a final volume of 300 μl. The reaction mixture was incubated at 37°C for 2, 4, 6, and 21 hours, and the culture supernatant was collected and centrifuged for ELISA. The cytokines IFN-γ, TNF-α, IGF-1, and IL-10 were measured using an ELISA kit.

[0110] The cytokine content of each sample measured by ELISA kit is shown in the table below. [Table 11] [Table 12] [Table 13] [Table 14]

[0111] Thus, during the process of target cell killing by PBMCs mediated by humanized bispecific antibodies, the cytokine secretion levels of PBMC cells induced by the anti-CD26-CD3 humanized bispecific antibodies 19G294 and 18G272 derived from murine antibody 62 were lower than that of 17G29 (an anti-CD26-CD3 humanized bispecific antibody derived from YS110) at all time points under the same dosage conditions. These molecular constructs that target CD3 and activate T cells are safer when the cytokine storm caused by the T cell cytokine secretion is relatively low (A FAD oncology analysis of CD3 bispecific constructs and first-in-human dose selection (2017), Regulatory Toxicology and Pharmacology, 90:144-152.). When compared under the same conditions with 17G29, which recognizes the same CD3 sequence, the bispecific antibodies 19G294 and 18G272 of the present application had the advantage of inducing lower cytokine secretion from T cells. This may be related to the fact that the epitope of the CD26 antigen recognized by their CDR regions is different from that of 17G29. Compared to other epitopes, the epitope of the antigen recognized by the antibodies of the present application is associated with a reduced risk of cytokine storm and improved safety.

[0112] Example 14 Humanized Bispecific Antibody-Mediated Toxicity to PBMCs 1, 786-0 cell model PBMC cells were labeled with the fluorescent dye Calcein-AM to produce a green fluorescent signal, and the cell concentration was 6 × 10 550 μl of 786-0 cells were seeded per well into a U-shaped 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of humanized bispecific antibody 18G272 was added to the corresponding sample reaction wells at final concentrations of 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of medium was added to the blank control wells. 50 μl of Triton X-100 was added to the positive control wells at a final concentration of 1%. Furthermore, 50 μl of Triton X-100 at a concentration of 9 × 10 at an E / T ratio of 10:1 was added to the positive control wells. 6 50 μl of PBMC cells (cells / ml) was added, and the reaction system was incubated at 37°C under carbon dioxide for 5 hours. After the reaction was completed, the supernatant was collected by centrifugation and placed in a new 96-well plate. The plate was then centrifuged again, and 80 μl of the supernatant was collected and placed in a black 96-well plate. The results were measured using a microplate reader under conditions of an excitation wavelength of 470 nm and an emission wavelength of 515 nm.

[0113] 2. OS-RC-2 cell model PBMC cells were labeled with the fluorescent dye Calcein-AM to produce a green fluorescent signal, and the cell concentration was 6 × 10 5 50 μl of OS-RC-2 cells were seeded per well into a U-shaped 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml. 50 μl of humanized bispecific antibodies were added to the corresponding sample reaction wells at final concentrations of 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml, respectively. 50 μl of medium was added to the blank control wells. 50 μl of Triton X-100 at a final concentration of 1% was added to the positive control wells. Furthermore, 9 × 10 mAb was added to the blank control wells at an E / T ratio of 15:1. 6 50 μl of PBMC cells (cells / ml) was added, and the reaction system was incubated at 37°C under carbon dioxide for 5 hours. After the reaction was completed, the supernatant was collected by centrifugation and placed in a new 96-well plate. The plate was then centrifuged again, and 80 μl of the supernatant was collected and placed in a black 96-well plate. The results were measured using a microplate reader under conditions of an excitation wavelength of 470 nm and an emission wavelength of 515 nm.

[0114] 3.Result analysis Cell lysis rate=(V sample -Vvehicle control ) / (V TritonX-100 -V vehicle control ) × 100%. (where V sample is the mean value of the fluorescent signal readings of the drug-treated group, and V vehicle control is the mean fluorescence signal reading of the solvent control group, and V TritonX-100 (The value is the average fluorescence signal reading of the positive control group.) The cell lysis rate and sample concentration values ​​were calculated using the software GraphPad Prism 7.00 to determine the IC value of each sample on PBMC cells. 50 The value was calculated.

[0115] During the process of antibody 18G272-mediated killing of target cells 786-0 by PBMC, the ability of PBMC to induce killing of PBMC cells was extremely low. The IC of antibody 18G272-mediated cytotoxic effect on PBMC(4#) cells was 50 The value was approximately 99,730,000 pg / ml, indicating that antibody 18G272 had no cytotoxic effect on PBMC(6#) cells. The results are shown in Figure 4.

[0116] The antibody 18G272-mediated killing of target cells OS-RC-2 by PBMC did not induce PBMC cell killing. Antibody 18G272 did not induce cytotoxicity of PBMC(1#) cells, nor did antibody 18G272 induce cytotoxicity of PBMC(3#) cells.

[0117] During the process of antibody 19G294-mediated killing of target cells 786-0 by PBMC, the ability of PBMC to cause killing of PBMC cells was extremely low. IC of antibody 19G294-mediated cytotoxic effect on PBMC(2#) cells 50 The IC value was approximately 1,915,313 pg / ml, indicating the IC value of the cytotoxic effect of antibody 19G294 on PBMC(3#) cells. 50 The value was approximately 1947224 pg / ml.

[0118] During the process of antibody 19G294-mediated killing of target cells OS-RC-2 by PBMC, the ability of PBMC to cause killing of PBMC cells was extremely low. IC of antibody 19G294-mediated cytotoxic effect on PBMC(2#) cells 50 The IC value was approximately 47,920,315 pg / ml, indicating the IC value of the cytotoxic effect of antibody 19G294 on PBMC(3#) cells. 50 The value was approximately 1406115138 pg / ml. As described above, the CD26-CD3 humanized bispecific antibodies can accurately recognize PBMCs. Although it has been reported in the literature that T cells express CD26, the bispecific antibodies 18G272 and 19G294 of the present application did not induce significant cytotoxic effects on PBMCs, demonstrating their high safety.

[0119] Example 15 Comparison of in vitro efficacy of 19G294 and drug B (BAVENCIO, PD-L1 antibody) Target cells were cultured in T75 cell culture flasks, and when the cells were more than 80% confluent, they were collected by trypsin digestion, washed once with PBS, and counted with a hemocytometer to obtain 5 × 10 cells. 5 The target cells were co-incubated with PE-labeled anti-PD-L1 monoclonal antibody (purchased from Sino Biological) for approximately 40 minutes at room temperature in the dark. After incubation, the cells were centrifuged, the supernatant was discarded, the cell pellet was resuspended in PBS, and the cells were centrifuged again, the supernatant was discarded, the cell pellet was collected, and the cell pellet was resuspended in approximately 200 μl of PBS solution. The PD-L1 positivity rate was measured and analyzed using a flow cytometer within 1 hour. Flow cytometry revealed that the PD-L1 positivity rate of the 786-0 cell line was 66.2%, while that of the A498 cells was 29.6%. [Table 15]

[0120] Human renal carcinoma 786-0-LUC cells (786-0 cells transfected with the reporter gene Luciference) and human renal carcinoma A498-LUC cells (A498 cells transfected with the reporter gene Luciference) were cultured at a cell density of 3 × 10 5 50 μl per well was seeded into a white-bottomed opaque 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml in final concentrations, respectively, into the corresponding reaction wells. 50 μl of antibody was added to the corresponding reaction wells at a final concentration of 3 × 10 cells / ml at an E / T ratio of 10:1. 6 50 μl of PBMC cells at 100 μg / ml was added, and the reaction system was incubated at 37°C under carbon dioxide for 21 hours. After the reaction was completed, the mixture was centrifuged, the supernatant was discarded, 100 μl of reporter gene assay reagent (purchased from Vazyme) was added, and the mixture was shaken at 400 rpm for 5 minutes on a microplate shaker. Immediately afterwards, measurements were performed using a microplate reader under Lum conditions. Using the measured data, a four-parameter curve was created based on the reporter gene reduction level and antibody amount using GraphPad Prism 7.00 software, and EC 50 The values ​​were calculated and the results are shown in Table 16. [Table 16]

[0121] The above data indicate that the EC values ​​of 19G294-mediated cytotoxicity in PBMCs against two types of target cells are 50 The PD-L1 expression rates of 786-0 and A498 were 66.2% and 29.6%, respectively, demonstrating that the activity of 19G294 is significantly higher than that of BAVENCIO, regardless of the PD-L1 expression level of the cells.

[0122] Example 16 Comparison of in vitro efficacy of 21G587 and 19G294 Human renal carcinoma 786-0-LUC cells (786-0 cells transfected with the reporter gene Luciference) and human renal carcinoma A498-LUC cells (A498 cells transfected with the reporter gene Luciference) were cultured at a cell density of 3 × 10 5 50 μl per well was seeded into a white-bottomed opaque 96-well cell culture plate at 100 ng / ml, 10 ng / ml, 1 ng / ml, 0.1 ng / ml, 0.01 ng / ml, or 0.001 ng / ml in final concentrations, respectively, into the corresponding reaction wells. 50 μl of antibody was added to the corresponding reaction wells at a final concentration of 3 × 10 cells / ml at an E / T ratio of 10:1. 6 PBMC cells (cells / ml) were added, and the reaction system was incubated at 37°C under carbon dioxide for 21 hours. After the reaction was completed, the mixture was centrifuged to discard the supernatant, and 100 μl of reporter gene assay reagent (purchased from Vazyme) was added. The mixture was shaken at 400 rpm on a microplate shaker for 5 minutes, and then immediately measured under Lum conditions using a microplate reader. Using the measured data, a four-parameter curve was created based on the reporter gene reduction level and antibody amount, and EC 50 The values ​​were calculated and the results are shown in Table 17. [Table 17]

[0123] Based on the above data, when the anti-CD26 antibody sequence of the present application is combined with a different anti-CD3 sequence, for example, 19G294 and 21G587 have the same anti-CD26 portion but different anti-CD3 portions, both have the same level of in vitro efficacy.

[0124] Example 17: Efficacy of humanized bispecific antibodies 18G272 and 19G294 in NOD / SCID mice bearing PC-3 (prostate cancer) xenograft tumors Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and cultured at 5 × 10 6 Cells / 0.1 ml of PC-3 cell suspension and 1 x 10 7The PBMC cell suspension (0.1 ml / 0.1 ml) was thoroughly mixed at a 1:1 volume ratio and inoculated subcutaneously into NOD / SCID mice. The inoculation volume per mouse was 0.2 ml. Animals were randomly divided into three groups based on body weight: the model group (PBS), the 19G294 group (30 μg / mouse / injection), and the 18G272 group (30 μg / mouse / injection). Each group consisted of six animals. The model group and each treatment group were administered intravenously starting on the day of inoculation. Administration was continued for five consecutive days, followed by a two-day rest period before the second course of administration. The first administration was administered one hour after subcutaneous inoculation of the cells. The first course was administered on days 1, 2, 3, 4, and 5, and the second course was administered on days 8, 9, 10, 11, and 12. The administration frequency was once daily. General clinical observations: During the quarantine and experimental periods, animals should be observed at least once daily for tumor growth and the effects of treatment on normal behavior, including death or moribundity of tumor-bearing animals, mental status, behavioral activity, and other abnormalities. Weight: Measure 2-3 times per week. Tumor volume: Measure 2-3 times weekly. Measure the long and short diameters of the tumor with a caliper. Tumor volume (mm 3 ) = major axis * minor axis 2 / 2. Relative tumor inhibition rate: TGI (%) = (1-T / C) x 100%. Generally, T represents the relative tumor volume at a certain time point in the administration group (the ratio of the tumor volume measured this time to the tumor volume at the time of grouping), and C represents the relative tumor volume at a certain time point in the model group (the ratio of the tumor volume measured this time to the tumor volume at the time of grouping). However, in this application, since the animals are divided into groups based on body weight on the day of inoculation in the experiment, when calculating TGI, T and C represent the tumor volumes actually measured this time in the administration group and the model group, respectively. Tumor weight: After the final measurement, the experimental animals were euthanized, the tumor masses were excised, washed with saline, and dried with filter paper. The tumor masses were then weighed and photographed. Relative tumor inhibition rate TGI (%) = (1-T TW / C TW )×100%, T TW represents the mean tumor weight of the treatment group at the end of the experiment, and C TWindicates the average tumor weight of the Model group at the end of the experiment. Conclusions: 18G272 at a dose of 30 μg / mouse tended to inhibit tumor growth in PC-3 human prostate cancer model mice, while 19G294 almost completely inhibited tumor growth in PC-3 human prostate cancer model mice. The results are shown in Figure 5 and Table 18. At a dose of 30 μg / mouse, test products 19G294 and 18G272 tended to increase animal weight, which was comparable to the weight of animals in the model group, and were well tolerated during the treatment period. The results are shown in Figure 6 and Table 18. [Table 18]

[0125] Example 18: Efficacy of humanized bispecific antibody 18G272 in NOD / SCID mice bearing NCI-H596 (lung cancer) xenograft tumors Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and cultured in 8 × 10 6 Cells / 0.1 ml of NCI-H596 cell suspension and 1.6 × 10 7 The PBMC cell suspension was thoroughly mixed at a 1:1 volume ratio of 18G272 cells / 0.1 ml and inoculated subcutaneously into NOD / SCID mice. The volume per mouse was 0.2 ml. The animals were randomly divided into two groups based on body weight: the model group (PBS, 6 mice / group) and the 18G272 group (30 μg / mouse / dose, 6 mice / group). The administration route and administration interval were the same as in Example 17. General clinical observations and methods for measuring body weight and tumor volume were the same as in Example 17. Conclusion: Test substance 18G272 significantly inhibited tumor growth in a mouse model of human lung adenocarcinoma at a dose of 30 μg / mouse. Test substance 18G272 tended to increase animal weight at a dose of 30 μg / mouse, and was well tolerated during treatment. The results are shown in Figure 7 and Table 19. [Table 19]

[0126] Example 19: Efficacy of humanized bispecific antibody 19G294 in NOD / SCID mice bearing NCI-H226 (mesothelioma) xenograft tumors Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and cultured at 5 × 10 6 Cells / 0.1 ml of NCI-H226 cell suspension and 1.5 × 10 7 The PBMC cell suspension (19G294 cells / 0.1 ml) was thoroughly mixed at a 1:1 (volume ratio) and subcutaneously inoculated into NOD / SCID mice. The inoculation volume per mouse was 0.2 ml. The animals were randomly divided into two groups based on body weight: the model group (PBS, 5 mice / group) and the 19G294 group (60 μg / mouse / dose, 5 mice / group). The administration route and interval were the same as in Example 17. General clinical observations and methods for measuring body weight and tumor volume were the same as in Example 17. Conclusion: Test product 19G294 significantly inhibited tumor growth in a mouse model of human pulmonary adenocarcinoma (mesothelioma) at a dose of 60 μg / mouse, as shown in Figure 8. Test product 19G294 at a dose of 60 μg / mouse tended to increase animal weight and was well tolerated during treatment. The results are shown in Figure 9 and Table 20. [Table 20]

[0127] Example 20: Efficacy of humanized bispecific antibody 19G294 in NOD / SCID mice bearing A498 (renal carcinoma) xenograft tumors Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and cultured at 1 × 10 7 Cells / 0.1 ml of A498 cell suspension and 1 x 10 7 The PBMC cell suspension was thoroughly mixed at a 1:1 volume ratio of 19G294 cells / 0.1 ml and inoculated subcutaneously into NOD / SCID mice. The inoculation volume per mouse was 0.2 ml. The animals were randomly divided into two groups based on body weight: the model group (PBS) and the 19G294 group (30 μg / mouse / injection). Each group consisted of five animals. The administration route and administration interval were the same as in Example 17. General clinical observations and methods for measuring body weight and tumor volume were the same as in Example 17. Conclusion: Test product 19G294 significantly inhibited tumor growth in human renal cancer model mice at a dose of 30 μg / mouse, as shown in Figure 10 and Table 21. Test product 19G294 did not cause weight loss in the animals at a dose of 30 μg / mouse, and was well tolerated during the treatment period. [Table 21]

[0128] Example 21: Efficacy of humanized bispecific antibodies 18G272, 19G294, etc. in NOD / SCID tumor-bearing mice with OS-RC-2 (renal carcinoma) xenograft tumors Test 1: Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and cultured at 5 × 10 6 Cells / 0.1 ml of OS-RC-2 cell suspension and 1 x 10 7 The PBMC cell suspension was thoroughly mixed at a volume ratio of 1:1 (2000 cells / 0.1 ml) and inoculated subcutaneously into NOD / SCID mice. The inoculation volume per mouse was 0.2 ml. The animals were randomly divided into three groups based on body weight: the model group (PBS), the 17G29 group (15 μg / mouse / injection), and the 19G295 group (15 μg / mouse / injection). Each group consisted of eight animals. The administration route and administration interval were the same as in Example 17. General clinical observations and methods for measuring body weight and tumor volume were the same as in Example 17. Conclusions: At a dose of 15 μg / mouse, both 17G29 and 19G295 significantly inhibited tumor growth at day 15 (Figure 11). However, at day 39, neither 17G29 nor 19G295 inhibited tumor growth; there was no significant difference between the two. Furthermore, although the OS-RC-2 human renal carcinoma model experienced some animal deaths during the experimental period, both 17G29 and 19G295 at a dose of 15 μg / mouse extended the median survival time of mice; there was no statistically significant difference between the two. [Table 22]

[0129] Test Two: Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and incubated with 3 × 10 6 Cells / 0.1 ml of OS-RC-2 cell suspension and 6 x 10 6 The PBMC cell suspension was thoroughly mixed at a 1:1 volume ratio of 18G272 cells / 0.1 ml and inoculated subcutaneously into NOD / SCID mice. The volume per mouse was 0.2 ml. The animals were randomly divided into three groups based on body weight: the model group (PBS, 6 mice), the 18G272 group (30 μg / mouse / dose, 6 mice), and the 19G295 group (30 μg / mouse / dose, 6 mice). The administration route and administration interval were the same as in Example 17. General clinical observations and methods for measuring body weight and tumor volume were the same as in Example 17. Conclusions: Compared to the model group, 18G272 significantly inhibited tumor growth at a dose of 30 μg / mouse at D22. Meanwhile, 19G295 tended to inhibit tumor growth, but the difference was not significant. At D30, the 18G272 group still had a good tumor-inhibiting effect, with a significant difference. Meanwhile, the 19G295 group tended to inhibit tumor growth, but the difference was not significant compared to the model group. The results are shown in Figure 12 and Table 23. While the OS-RC-2 human renal carcinoma model experienced animal deaths during the experimental period, the test product 18G272 did not cause any animal deaths at a dose of 30 μg / mouse. No animal deaths occurred in the 19G295 group. These results are shown in Figure 13 and Table 23. Thus, 18G272 has a superior antitumor effect to 19G295. [Table 23]

[0130] Test Three: Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and incubated with 3 × 10 6 Cells / 0.1 ml of OS-RC-2 cell suspension and 6 x 10 6The PBMC cell suspension was thoroughly mixed at a volume ratio of 1:1 (2000 cells / 0.1 ml) and subcutaneously inoculated into NOD / SCID mice. The volume per mouse was 0.2 ml. The animals were randomly divided into two groups based on body weight: the model group (PBS) and the 18G272 group (30 μg / mouse / injection). Each group consisted of six animals. The administration route and administration interval were the same as in Example 17. General clinical observations, body weight, tumor volume, and tumor weight were measured in the same manner as in Example 17. Conclusion: 18G272 at a dose of 30 μg / mouse significantly inhibited tumor growth in OS-RC-2 human renal carcinoma model mice. The results are shown in Figure 14 and Table 24. Although animal deaths occurred during the experimental period in the OS-RC-2 human renal carcinoma model, under the conditions of this experiment, 18G272 at a dose of 30 μg / mouse significantly extended the median survival time (MST) of the OS-RC-2 human renal carcinoma model mice, reduced animal mortality, and was well tolerated during the treatment period. The results are shown in Table 24. [Table 24]

[0131] Test 4: Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and incubated with 3 × 10 6 Cells / 0.1 ml of OS-RC-2 cell suspension and 6 x 10 6 The PBMC cell suspension was thoroughly mixed at a 1:1 volume ratio of 19G294 cells / 0.1 ml and inoculated subcutaneously into NOD / SCID mice. The inoculation volume per mouse was 0.2 ml. The animals were randomly divided into two groups based on body weight: the model group (PBS) and the 19G294 group (30 μg / mouse / injection). Each group consisted of five animals. The administration route and administration interval were the same as in Example 17. General clinical observations and methods for measuring body weight and tumor volume were the same as in Example 17. Conclusion: 19G294 significantly inhibited tumor growth in the OS-RC-2 human renal carcinoma model mice at a dose of 30 μg / mouse. The results are shown in Figure 15 and Table 25. Although the OS-RC-2 human renal carcinoma model resulted in animal deaths during the experimental period, under the conditions of this experiment, 19G294 reduced animal mortality, prolonged animal survival, and was well tolerated during the treatment period. The results are shown in Table 25. [Table 25]

[0132] Taking the above four experimental results into account, in the OS-RC-2 (renal carcinoma) xenograft tumor-bearing NOD / SCID mouse model, 17G29 was comparable to 19G295 in terms of efficacy, 19G295 was inferior to 18G272, and 19G294 was comparable to 18G272, with both significantly or completely inhibiting tumor growth. In terms of safety, 19G295 was comparable to 17G29, 19G295 was inferior to 18G272, and 19G294 was comparable to 18G272. In summary, 18G272 and 19G294 have superior in vivo efficacy and safety to 17G29.

[0133] Example 22: Efficacy of different doses of humanized bispecific antibodies 19G294 and 21G587 in NOD / SCID mice bearing OS-RC-2 (renal carcinoma) xenograft tumors Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and incubated with 3 × 10 6 Cells / 0.1 ml of OS-RC-2 cell suspension and 6 x 10 6 The PBMC cell suspension was thoroughly mixed at a 1:1 volume ratio of 19G294 cells / 0.1 ml and inoculated subcutaneously into NOD / SCID mice. The volume per mouse was 0.2 ml. The animals were randomly divided into five groups based on body weight: Model group (PBS), 19G294 group (5 μg / mouse / injection), 19G294 group (15 μg / mouse / injection), 19G294 group (30 μg / mouse / injection), and 21G587 group (30 μg / mouse / injection). Each group consisted of three animals. The administration route and administration interval were the same as in Example 17. General clinical observations, body weight, tumor volume, and tumor weight were measured in the same manner as in Example 17. Conclusions: Test compounds 19G294 at doses of 15 and 30 μg / mouse and 21G587 at a dose of 30 μg / mouse significantly inhibited tumor growth in the OS-RC-2 human renal carcinoma mouse model. The efficacy of 19G294 was dose-dependent at low, medium, and high doses. Although animal deaths occurred during the experimental period in the OS-RC-2 human renal carcinoma model, under the experimental conditions, test compounds 19G294 (5, 15, and 30 μg / mouse) and 21G587 (30 μg / mouse) ameliorated the decline in animal health due to tumor growth, reduced animal mortality, and were well tolerated during the treatment period. The results are shown in Figure 16, Figure 17, and Table 26. [Table 26]

[0134] Example 23: Efficacy of humanized bispecific antibody 19G294, Axitinib + BAVENCIO combination therapy, and YS110 in NOD / SCID tumor-bearing mice with OS-RC-2 (renal carcinoma) xenograft tumors Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and incubated with 3 × 10 6 Cells / 0.1 ml of OS-RC-2 cell suspension and 6 x 10 6The PBMC cell suspension (0.1 ml / 0.1 ml) was thoroughly mixed at a 1:1 (volume ratio) and inoculated subcutaneously into NOD / SCID mice. The volume per mouse was 0.2 ml. Animals were randomly divided into five groups based on body weight: model group (PBS), positive control group A (BAVENCIO, 1.8 mg / mouse / dose; Axitinib, 30 μg / mouse / dose), positive control group B (BAVENCIO, 1.8 mg / mouse / dose; Axitinib, 60 μg / mouse / dose), 19G294 group (30 μg / mouse / dose), and YS110 group (205 μg / mouse / dose). Each group consisted of three to four animals. Axitinib was administered intragastrically, while BAVENCIO, the test product, and PBS were administered intravenously. Axitinib, 19G294, and PBS will be administered starting on the day of vaccination for five consecutive days, followed by a two-day rest period before the second course (five consecutive days), with a daily dosing frequency of 10 doses in total. BAVENCIO will be administered starting on the day of vaccination, once weekly for two consecutive weeks (two doses in total). YS110 will be administered starting on the day of vaccination, twice weekly for two consecutive weeks (four doses in total). General clinical observations, body weight, tumor volume, and tumor weight were measured in the same manner as in Example 17. Conclusions: Test compound 19G294, at a dose of 30 μg / mouse, significantly inhibited tumor growth in the OS-RC-2 human renal carcinoma mouse model. Neither BAVENCIO (1.8 mg / mouse) + Axitinib (30 μg / mouse, 60 μg / mouse) combination therapy nor YS110 (205 μg / mouse) significantly inhibited tumor growth in the OS-RC-2 human renal carcinoma mouse model. The OS-RC-2 human renal carcinoma model is prone to significant weight loss or death during the experimental period. Under the conditions of this study, test compound 19G294 (30 μg / mouse) ameliorated the decline in animal health due to tumor growth, reduced animal mortality, and was well tolerated during the treatment period. Neither BAVENCIO + Axitinib combination (both dose groups) nor YS110 produced these effects. In summary, under the conditions of this experiment, the efficacy of 19G294 was superior to that of BAVENCIO + Axitinib combination and YS110. The results are shown in Figure 18, Figure 19 and Table 27. [Table 27]

[0135] Example 24: Efficacy of humanized bispecific antibody 19G294, Axitinib + BAVENCIO combination therapy, and YS110 in NOD / SCID tumor-bearing mice with A498 (renal carcinoma) xenograft tumors Methods: NOD / SCID mice weighing 18-22 g and approximately 5-7 weeks old were selected and cultured at 7 × 10 6 Cells / 0.1 ml of A498 cell suspension and 7 x 10 6 The PBMC cell suspension (0.1 ml / 0.1 ml) was thoroughly mixed at a 1:1 (volume ratio) and inoculated subcutaneously into NOD / SCID mice. The volume per mouse was 0.2 ml. Animals were randomly divided into five groups based on body weight: model group (PBS), positive control group B (BAVENCIO, 1.8 mg / mouse / dose; Axitinib, 60 μg / mouse / dose), positive control group C (BAVENCIO, 1.8 mg / mouse / dose; Axitinib, 120 μg / mouse / dose), 19G294 group (30 μg / mouse / dose), and YS110 group (205 μg / mouse / dose). Each group consisted of three animals. Axitinib was administered intragastrically, while BAVENCIO, the test product, and PBS were administered intravenously. Axitinib, 19G294, and PBS were administered starting on the day of vaccination for five consecutive days, followed by a two-day rest period before the second course (five consecutive days), with a daily dosing frequency of 10 doses in total. BAVENCIO was administered starting on the day of vaccination, once every two weeks (one dose in total). YS110 was administered starting on the day of vaccination, twice weekly for two consecutive weeks (four doses in total). General clinical observations, body weight, tumor volume, and tumor weight were measured in the same manner as in Example 17. Organ weight and organ index: After the final measurement, the animals were euthanized, and the organs were separated, rinsed with saline, blotted dry with filter paper, and weighed and photographed. Organ index = organ weight / mouse body weight × 100% (organ weight and mouse body weight are in grams). Conclusions: Test compound 19G294, at a dose of 30 μg / mouse, significantly inhibited tumor growth in A498 human renal carcinoma mouse models. Neither BAVENCIO (1.8 mg / mouse) + Axitinib (60 μg / mouse, 120 μg / mouse) combination therapy nor YS110 (205 μg / mouse) significantly inhibited tumor growth in OS-RC-2 human renal carcinoma mouse models. Test compound 19G294 (30 μg / mouse), BAVENCIO + Axitinib combination (two dose groups), nor YS110 were well tolerated during treatment. Test compound 19G294 (30 μg / mouse) nor YS110 significantly affected the kidney, spleen, liver, lung, or corresponding organ parameters in mice. The combination of BAVENCIO and Axitinib (both dose groups) had significant effects on the liver and / or corresponding organ parameters, but no significant effects on the kidney, spleen, lung, and corresponding organ parameters. In summary, under the conditions of this experiment, the efficacy of 19G294 was superior to that of the combination of BAVENCIO and Axitinib and YS110, and its safety was superior to that of the combination of BAVENCIO and Axitinib. The results are shown in Figure 20, Figure 21, Table 28, and Table 29. [Table 28] [Table 29-1] [Table 29-2]

Claims

1. An antibody or antigen-binding fragment that specifically binds to human CD26, characterized in that it contains an HCDR1 shown in SEQ ID NO: 1, an HCDR2 shown in SEQ ID NO: 2, an HCDR3 shown in SEQ ID NO: 3, and an LCDR1 shown in SEQ ID NO: 4, an LCDR2 consisting of Arg Met Ser, and an LCDR3 shown in SEQ ID NO:

5.

2. 2. The antibody or antigen-binding fragment of claim 1, comprising a heavy chain variable region whose amino acid sequence is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to the sequence set forth in SEQ ID NO:6 or SEQ ID NO:8, and a light chain variable region whose amino acid sequence is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to the sequence set forth in SEQ ID NO:7 or SEQ ID NO:

9.

3. The antibody or antigen-binding fragment of claim 2, characterized in that 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acids are inserted, deleted, or substituted in the sequence set forth in SEQ ID NO: 6, SEQ ID NO: 7, SEQ ID NO: 8, or SEQ ID NO: 9, and the amino acid substitutions are conservative substitutions of Ala to Gly or Ser, Arg to Lys or His, or Asn to Gln or His.

4. An antibody or antigen-binding fragment according to claim 2, characterized in that it comprises a heavy chain variable region shown in SEQ ID NO: 6 and a light chain variable region shown in SEQ ID NO: 7, or a heavy chain variable region shown in SEQ ID NO: 8 and a light chain variable region shown in SEQ ID NO:

9.

5. a bispecific T cell-inducing antibody comprising a sequence of a CD26 antibody or antigen-binding fragment thereof, the bispecific T cell-inducing antibody being formed by linking two single-chain variable region fragments (scFv) having antigen specificity with a linker, one scFv recognizing CD26 as a target and formed by linking an antibody heavy chain variable region recognizing CD26 with an antibody light chain variable region with the linker, and the other scFv recognizing CD3 as a target and formed by linking an antibody heavy chain variable region recognizing CD3 with an antibody light chain variable region with the linker, the scFv recognizing CD26 as a target comprising an HCDR1 shown in SEQ ID NO: 1, an HCDR2 shown in SEQ ID NO: 2, an HCDR3 shown in SEQ ID NO: 3, and an LCDR1 shown in SEQ ID NO: 4, an LCDR2 consisting of ArgMetSer, and an LCDR3 shown in SEQ ID NO:

5.

6. 6. A bispecific T cell-inducing antibody (BiTE) comprising the sequence of the CD26 antibody or antigen-binding fragment according to claim 5, wherein the scFv that recognizes CD26 as a target comprises a heavy chain variable region whose amino acid sequence is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to the sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 8, and a light chain variable region whose amino acid sequence is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to the sequence set forth in SEQ ID NO: 7 or SEQ ID NO:

9.

7. A bispecific T cell-inducing antibody (BiTE) comprising the sequence of the CD26 antibody or antigen-binding fragment described in claim 6, characterized in that the scFv that recognizes CD26 as a target comprises a heavy chain variable region shown in SEQ ID NO: 6 and a light chain variable region shown in SEQ ID NO: 7, or a heavy chain variable region shown in SEQ ID NO: 8 and a light chain variable region shown in SEQ ID NO:

9.

8. A bispecific T cell-inducing antibody (BiTE) comprising the sequence of the CD26 antibody or antigen-binding fragment described in claim 5, characterized in that the scFv that recognizes CD3 as a target is derived from OKT-3, L2K, TR66, UCHT1, SP34, IORT3, Catumaxomab, Blinatumomab, or Solitomab.

9. 6. A bispecific T cell-inducing antibody (BiTE) comprising the sequence of the CD26 antibody or antigen-binding fragment according to claim 5, wherein the linker connecting the heavy chain variable region and the light chain variable region of the scFv is selected from the group consisting of KESGSVSSEQLAQFRSLD, EGKSSGSGSESKST, GSTSGGGSGGGSGGGGSS, GSTSGSGKPGSGEGSTKG, and (GGGGS). n and n is an integer from 1 to 5.

10. A bispecific T cell-inducing antibody comprising the sequence of the CD26 antibody or antigen-binding fragment described in claim 9, characterized in that the sequence of the linker connecting the heavy chain variable region and light chain variable region of the scFv is shown in SEQ ID NO:

12.

11. 6. A bispecific T cell-inducing antibody comprising the sequence of the CD26 antibody or antigen-binding fragment of claim 5, wherein the linker connecting the two scFvs is (GGGGS) n , n is an integer from 1 to 5.

12. A bispecific T cell-inducing antibody comprising the sequence of the CD26 antibody or antigen-binding fragment described in claim 5, characterized in that the sequence of the linker connecting the two scFvs is selected from the sequence shown in SEQ ID NO:

13.

13. A bispecific T cell-inducing antibody comprising the sequence of the CD26 antibody or antigen-binding fragment of claim 5, wherein the amino acid sequence of the bispecific T cell-inducing antibody is set forth in SEQ ID NO: 14 or SEQ ID NO:

15.

14. A polynucleotide encoding the amino acid sequence of the antibody or antigen-binding fragment of any one of claims 1-13.

15. A vector comprising the polynucleotide of claim 14.

16. A host cell comprising the vector of claim 15.

17. A pharmaceutical composition comprising the antibody or antigen-binding fragment of any one of claims 1 to 13.

Images