Bispecific antibody targeting SIRPα and PD-L1 or antigen-binding fragment thereof and applications

Abstract

Bispecific antibodies or antigen-binding fragments thereof targeting SIRPα and PD-L1 and their applications. The bispecific antibody comprises an SIRPα-binding domain and a PD-L1-binding domain. The SIRPα-binding domain comprises a heavy-chain variable region and a light-chain variable region, and the PD-L1-binding domain comprises a VHH fragment. Further provided are drugs comprising the bispecific antibody or antigen-binding fragment thereof targeting SIRPα and PD-L1, as well as nucleic acid molecules, vectors, host cells transformed with the vectors, and pharmaceutical uses of the antibodies.

Description

Detailed Description of the Invention

[0001] [Technical Field] The present invention relates to the field of biomedical technology, and specifically, to bispecific antibodies or antigen-binding fragments thereof targeting SIRPα and PD-L1, and their applications.

[0002] [Background Art] PD-1 (CD279) was first reported in 1992. The human PD-1 coding gene, PDCD1, is located at 2q37.3, with a full length of 2097 bp and is composed of six exons. PD-1 is a membrane protein belonging to the CD28 immunoglobulin superfamily, mainly expressed on the surface of activated T cells, and also expressed in small amounts on CD4-CD8-T cells in the thymus, activated NK cells and monocytes. PD-1 has two ligands, namely PD-L1 (CD274, B7-H1) and PD-L2 (CD273, B7-DC) of the B7 protein family. PD-L1 and PD-L2 have 40% identical amino acid sequences. The main difference between the two lies in their expression patterns. PD-L1 is constitutively low-expressed in APCs, non-hematopoietic cells (such as vascular endothelial cells, pancreatic islet cells) and immune-privileged sites (such as placenta, testis and eye). Inflammatory cytokines such as type I and type II interferons, TNF-α and VEGF may induce the expression of PD-L1. PD-L2 is only expressed in activated macrophages and dendritic cells. After PD-1 and PD-L1 bind to activated T cells, the ITSM motif of PD-1 undergoes tyrosine phosphorylation, thereby dephosphorylating downstream protein kinases Syk and PI3K, inhibiting the activation of downstream pathways such as AKT and ERK, and ultimately inhibiting the transcription and translation of genes and cytokines necessary for T cell activation, playing a role in negatively regulating T cell activity. In tumor cells, tumor cells and the tumor microenvironment upregulate PD-L1 expression and bind to PD-1 on the surface of tumor-specific CD8+ T cells, thereby negatively regulating T cell activity and inhibiting the immune response. There is increasing evidence that tumors utilize PD-1-dependent immune inhibition for immune evasion. High expression of PD-L1 and PD-L2 has been observed in various solid tumors and hematological malignancies. Furthermore, there is a strong correlation between the expression of PD-Ls and poor prognosis of tumor cells.

[0003] Phagocytosis of tumor-associated macrophages (TAM) in the tumor microenvironment is inhibited because the CD47 protein is highly expressed on the surface of almost all tumor cells, binds to signal regulatory protein α (SIRPα) on the surface of myeloid cells, and sends a "don't eat me" or "self" signal to the body, thereby inhibiting phagocytosis. CD47, also known as integrin-associated protein (IAP), is a widely expressed transmembrane glycoprotein belonging to the immunoglobulin (Ig) superfamily. The molecular weight of CD47 is 50 kD, and its structure includes a large number of glycosylated N-terminal IgV variable domains, five highly hydrophobic transmembrane domains, and one short C-terminal cytoplasmic tail. The expression of CD47 in various tissues is determined by four alternative splicing forms of the C-terminal cytoplasmic tail. The corresponding SIRPα, also called SHPS-1, BIT, or CD172a protein, is a transmembrane protein mainly expressed in myeloid cells, including macrophages, bone marrow dendritic cells, granulocytes, mast cells, and their precursor cells. SIRPα is composed of three extracellular Ig-like domains and four tyrosine residues in the cytoplasm, and the four tyrosine residues are presumed to be phosphorylation sites. After phosphorylation, SIRPα binds to the SH2 domain of the SHP-1 / 2 protein and is activated, thereby activating the downstream signaling pathway. Since the expression of SHP-1 and SHP-2 proteins is tissue-specific, SIRPα is a docking protein that mobilizes and activates downstream protein phosphatases in response to extracellular stimuli. Oldenborg first reported that mature red blood cells (RBC) protect themselves from removal by the spleen macrophage SIRPα by binding through CD47.Subsequently, the researchers discovered that SIRPα on RBCs can also bind to monocyte SIRPα and inhibit Fcγ receptor-dependent phagocytosis, which is achieved by dephosphorylating myosin-IIA, an important molecule in phagocytosis. Clinically, high expression of CD47 has been observed in various solid tumors and hematological malignancies, including acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), chronic myeloid leukemia (CML), non-Hodgkin lymphoma (NHL), breast cancer, bladder cancer, ovarian cancer, colon cancer, etc. The essence is that tumor cells avoid the cell clearance effect of macrophages through the above regulatory mechanism. CD47 also affects other biological processes through binding to other receptors or signal transduction in its intracellular cytoplasmic region. The interaction between CD47 and thrombospondin-1 (TSP-1) or vascular endothelial growth factor receptor 2 (VEGFR-2) inhibits angiogenesis and thereby limits tumor growth.

[0004] Due to the biological function of CD47 itself, it has been determined that CD47 therapeutic antibodies and SIRPα-Fc recombinant proteins may have the risk of causing hematological toxicity or anemia, which has been reported in CD47 gene knockout NOD mice and mouse models treated with CD47 antibodies. Furthermore, it has been reported that endothelial cell CD47 interacts with SIRPγ through cell adhesion and promotes the transendothelial migration of T cells, but SIRPγ is expressed mainly in T cells rather than myeloid cells. Therefore, using SIRPα antibodies is a more preferable choice for blocking the CD47-SIRPα signaling pathway. In addition, the Weissman research group at Stanford University demonstrated that the combination of the screened humanized SIRPα antibody KWAR23 and rituximab can effectively inhibit the growth of Burkitt lymphoma in SRG mice (Rag2- / -Il2rγ- / -) knocked in with the human SIRPα gene, but there is no obvious drug effect when KWAR23 is used alone.

[0005] [Summary of the Invention] [Means for Solving the Problems] A first object of the present invention is to provide a bispecific antibody targeting SIRPα and PD-L1 or an antigen-binding fragment thereof. The bispecific antibody targeting SIRPα and PD-L1 or an antigen-binding fragment thereof provided by the present invention includes an SIRPα-binding domain and a PD-L1-binding domain, wherein the SIRPα-binding domain includes a heavy chain variable region and a light chain variable region. The heavy chain variable region includes VHCDR1, VHCDR2, and VHCDR3 whose amino acid sequences are shown in SEQ ID NOs: 3, 4, and 5 respectively, and the light chain variable region includes VLCDR1, VLCDR2, and VLCDR3 whose amino acid sequences are shown in SEQ ID NOs: 37, 38, and 9 respectively. The PD-L1-binding domain includes a VHH fragment, and the VHH fragment includes CDR1, CDR2, and CDR3 whose amino acid sequences are shown in SEQ ID NOs: 63, 64, and 65 respectively.

[0006] Optionally, the sequence of the heavy chain variable region of the SIRPα-binding domain is as shown in SEQ ID NO: 17, or has at least 85% sequence identity therewith, or the sequence of the light chain variable region of the SIRPα-binding domain is selected from SEQ ID NO: 18, or has at least 85% sequence identity therewith.

[0007] Optionally, the sequence of the VHH fragment is as shown in SEQ ID NO: 62, or has at least 85% sequence identity therewith.

[0008] Optionally, the bispecific antibody or an antigen-binding fragment thereof further includes a heavy chain constant region selected from human IgG1, IgG2, IgG3, or IgG4 or a variant thereof, and a light chain constant region selected from human κ chain, λ chain, or a variant thereof.

[0009] Optionally, the heavy chain constant region includes an Fc fragment or a variant thereof, and the variant of the Fc fragment is derived from IgG1 and contains mutation sites: L234A, L235A, and K338A according to the EU numbering.

[0010] Optionally, the bispecific antibody or an antigen-binding fragment thereof includes a first polypeptide chain and a second polypeptide chain. The first polypeptide chain includes the heavy chain variable region of the SIRPα-binding domain, the heavy chain constant region, and the VHH fragment. The VHH fragment is fused to the N-terminus of the heavy chain variable region of the SIRPα-binding domain or the VHH fragment is fused to the C-terminus of the heavy chain constant region. The second polypeptide chain includes the light chain variable region of the SIRPα-binding domain and the light chain constant region.

[0011] Optionally, the bispecific antibody or an antigen-binding fragment thereof includes a first polypeptide chain and a second polypeptide chain. The first polypeptide chain includes the heavy chain variable region and the heavy chain constant region of the SIRPα-binding domain. The second polypeptide chain includes the light chain variable region of the SIRPα-binding domain, the light chain constant region, and the VHH fragment. The VHH fragment is fused to the N-terminus of the light chain variable region of the SIRPα-binding domain.

[0012] Optionally, the bispecific antibody or an antigen-binding fragment thereof has a symmetric structure including two of the first polypeptide chains and two of the second polypeptide chains.

[0013] Optionally, the bispecific antibody or an antigen-binding fragment thereof further includes a linker sequence, and the linker sequence may be (GGGGS)n, where n is an integer from 1 to 4. Optionally, the amino acid sequence of the first polypeptide chain is as shown in any of SEQ ID NO: 66, 26, 69, 84, 85, or the amino acid sequence of the second polypeptide chain is as shown in any of SEQ ID NO: 67, 68, 82, 83.

[0014] Optionally, the amino acid sequence of the first polypeptide chain is as shown in SEQ ID NO: 66, and the amino acid sequence of the second polypeptide chain is as shown in SEQ ID NO: 67.

[0015] A second object of the present invention is to provide a drug comprising a bispecific antibody targeting the above SIRPα and PD-L1 or an antigen-binding fragment thereof.

[0016] Optionally, the drug further comprises one or more other cancer therapeutics. A third object of the present invention is to provide a nucleic acid molecule encoding a bispecific antibody targeting the above SIRPα and PD-L1 or an antigen-binding fragment thereof.

[0017] A fourth object of the present invention is to provide a vector comprising the above nucleic acid molecule. A fifth object of the present invention is to provide a host cell transformed with the above vector.

[0018] A sixth object of the present invention is to provide the use of a bispecific antibody targeting the above SIRPα and PD-L1 or an antigen-binding fragment thereof in the preparation of a drug for inhibiting or treating a disease, disorder or condition.

[0019] Optionally, the disease, disorder or condition includes cancer, solid tumor, chronic infection, inflammatory disease, multiple sclerosis, autoimmune disease, neurological disorder, brain injury, nerve injury, polycythemia, hemochromatosis, trauma, septic shock, fibrosis, atherosclerosis, obesity, type II diabetes, graft dysfunction or arthritis.

[0020] Optionally, the cancer is selected from anal cancer, appendiceal cancer, astrocytoma, basal cell cancer, gallbladder cancer, gastric cancer, lung cancer, bronchial cancer, bone cancer, hepatobiliary duct cancer, pancreatic cancer, breast cancer, liver cancer, ovarian cancer, testicular cancer, kidney cancer, renal pelvis and ureter cancer, salivary gland cancer, small intestine cancer, urethral cancer, bladder cancer, head and neck cancer, spinal cancer, brain cancer, cervical cancer, uterine cancer, endometrial cancer, colon cancer, colorectal cancer, rectal cancer, esophageal cancer, gastrointestinal cancer, skin cancer, prostate cancer, pituitary cancer, vaginal cancer, thyroid cancer, laryngeal cancer, glioblastoma, melanoma, myelodysplastic syndrome, sarcoma, teratoma, chronic lymphocytic leukemia (CLL), chronic myelogenous leukemia (CML), acute lymphoblastic leukemia (ALL), acute myelogenous leukemia (AML), Hodgkin lymphoma, non-Hodgkin lymphoma, multiple myeloma, T-cell or B-cell lymphoma, gastrointestinal stromal tumor, soft tissue tumor, hepatocellular cancer or adenocarcinoma.

[0021] Optionally, the drug is used in combination with one or more other drugs. Optionally, the other drug includes rituximab. [Advantages of the Invention] Compared with the prior art, the present invention has at least the following beneficial effects.

[0022] (1) The bispecific antibody provided by the present invention can simultaneously target SIRPα and PD-L1, mediate the killing of immune cells while targeting tumor cells, and avoid the risk of hematological toxicity or anemia by targeting SIRPα.

[0023] (2) The SIRPα binding domain sequence of the bispecific antibody of the present invention is novel. (3) The bispecific antibody provided by the present invention has a unique structure, can target target proteins with high efficiency, and can obtain an efficient tumor killing effect.

[0024] (4) The bispecific antibody provided by the present invention can bind to all subtypes of human SIRPα protein, which is beneficial for clinical development.

Brief Description of the Drawings

[0025]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Figure 24

Figure 25

Figure 26

Figure 27

Figure 28

Figure 29

Figure 30

Figure 31

Figure 32

Figure 33

Figure 34

Figure 35

Figure 36

Figure 37

Figure 38

Figure 39

Figure 40

Figure 41

Figure 42

Figure 43

Figure 44

Figure 45

Figure 46

Figure 47

Figure 48

Figure 49

Figure 50

Figure 51

Figure 52

Figure 53

Figure 54

Figure 55

Figure 56

Figure 57

Figure 58

Figure 59

Figure 60

Figure 61

Figure 62

Figure 63

Figure 64

Figure 65

Figure 66

Figure 67

Figure 68

Figure 69

Mode for Carrying Out the Invention

[0026] Terms: "Antibody" (Ab) refers to an immunoglobulin molecule (Ig) containing at least one antigen-binding site and capable of specifically binding to an antigen.

[0027] "Antigen" is a substance that induces an immune response in the body and specifically binds to an antibody. The binding between an antibody and an antigen is mediated by the interaction formed between the two, including hydrogen bonds, van der Waals forces, ionic bonds, and hydrophobic bonds. The binding region between the antigen surface and the antibody is the "antigen determinant" or "epitope", and generally, each antigen has multiple determinants.

[0028] "Fusion" refers to directly connecting components by peptide bonds or connecting components via one or more peptide linkers. Among various components of an antibody molecule, they are connected by a "peptide linker", which ensures the correct folding of the protein and the stability of the peptide. For the "peptide linker", an amino acid sequence with low immunogenicity can be selected. In this specification, "peptide linker" and "linker sequence" have the same meaning. The linker sequence connects each component part of the fusion protein, and when specifically implemented, appropriate linker sequences such as (GS)n, (GSGGS (SEQ ID NO: 87))n, (GGGS (SEQ ID NO: 88))n, (GGGGS (SEQ ID NO: 89))n, etc. can be selected. n can be selected from the numbers 1 to 4, or a number greater than or equal to 4.

[0029] The term "antibody" referred to in the present invention is understood in its broadest sense and includes monoclonal antibodies (including full-length monoclonal antibodies), polyclonal antibodies, antibody fragments, multispecific antibodies (e.g., bispecific antibodies) containing at least two different antigen-binding domains. Antibodies further include mouse antibodies, humanized antibodies, chimeric antibodies, human antibodies, and antibodies from other sources. The antibodies of the present invention can be derived from any animal, including but not limited to immunoglobulins of humans, non-human primates, mice, rats, cows, horses, chickens, camels, alpacas, etc. Antibodies may contain additional changes such as non-natural amino acids, mutations in Fc effector functions, and mutations in glycosylation sites. Antibodies further include post-translationally modified antibodies, fusion proteins containing antigenic determinants of antibodies, and immunoglobulin molecules containing any other modifications to the antigen recognition site, as long as these antibodies exhibit the desired biological activity.

[0030] The basic structure of a conventional antibody is a Y-shaped monomer in which two completely identical heavy chains (H) and two completely identical light chains (L) are linked by disulfide bonds. Each chain is composed of 2-5 domains (functional regions) with similar but different functions, each containing approximately 110 amino acids. In an antibody molecule, the amino acid sequences near the N-terminus of the light and heavy chains vary greatly, and the formed domains are called variable regions (V regions), while the regions where the amino acid sequences near the C-terminus are relatively constant are called constant regions (C regions).

[0031] The V regions of the heavy and light chains are called VH and VL, respectively. VH and VL each have the amino acid composition of three regions, and their sequence order is very easily variable and is called the hypervariable region (HVR). This region forms a spatial structure complementary to the antigen epitope and is also called the complementarity determining region (CDR). The three CDRs of VH are represented by VHCDR1, VHCDR2, and VHCDR3, respectively, and the three CDRs of VL are represented by VLCDR1, VLCDR2, and VLCDR3, respectively. The total six CDRs of VH and VL together form the antigen-binding site. The amino acid diversity in the CDR region is the molecular basis for antibodies to specifically bind to a large number of different antigens. The amino acid composition and arrangement order outside the CDRs of the V region are relatively less variable and are called the framework region or framework region (FR). VH and VL have four framework regions represented by FR1, FR2, FR3, and FR4. VH and VL are each composed of three CDRs and four FRs, and the sequence order from the amino group terminus (N-terminus) to the carboxyl group terminus (C-terminus) is FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4.

[0032] According to the amino acid sequence of the constant region of the antibody heavy chain, human immunoglobulins can be classified into five categories: IgM, IgG, IgA, IgD, and IgE. They can be further divided into different subtypes (isotypes). For example, human IgG can be divided into IgG1, IgG2, IgG3, and IgG4, and IgA can be divided into IgA1 and IgA2. Subtypes of IgM, IgD, and IgE have not been discovered. Based on the light chain amino acid sequence, the light chain can be classified into κ chain and λ chain. The antibody of the present invention can be of any type (e.g., IgM, IgG, IgA, IgD, IgE) or subtype (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, IgA2).

[0033] The constant regions of the heavy chain and the light chain are called CH and CL, respectively. The heavy chain constant regions of IgG, IgA, and IgD have three domains: CH1, CH2, and CH3, while the heavy chain constant regions of IgM and IgE have four domains: CH1, CH2, CH3, and CH4.

[0034] The region between CH1 and CH2 is the hinge region. Since it contains abundant proline, it is easy to stretch or bend, can change the distance between the two arms of the Y shape, and is advantageous for the two arms to bind to the antigen epitope simultaneously.

[0035] "Antigen-binding fragment" refers to Fab fragment, F(ab’)2 fragment, Fv fragment, ScFv fragment, etc. that have antigen-binding activity. "Fab fragment" (fragment of antigen binding, Fab) refers to an antibody fragment composed of VL, VH, CL, and CH1 domains that binds to a single antigen epitope (monovalent). Those skilled in the art know that papain hydrolyzes IgG to form two identical Fab segments and one Fc segment, and pepsin hydrolyzes IgG to form one F(ab’)2 segment and some polypeptide fragments (pFc’). When the disulfide bond between the F(ab’)2 heavy chains is cleaved, two Fab’ fragments can be formed, and the latter can be further enzymatically cleaved into Fv fragments. The Fv fragment contains the variable regions of the heavy and light chains of the antibody but does not contain the constant regions. A single-chain variable fragment scFv (single chain antibody fragment), or single-chain antibody, is one in which the variable regions of the heavy and light chains of the antibody are linked via a linker.

[0036] In 1993, the Hamers laboratory discovered that, in addition to conventional quadruple antibodies, camel serum also contains a large number of molecules similar to immunoglobulin G (IgG). Such molecules are called heavy-chain antibodies (HCAb). The constant regions CH1 of the light and heavy chains of conventional antibodies are naturally absent, but they still have a strong binding ability to antigens. The Hamers laboratory analyzed and identified the structure and sequence of the heavy-chain antibodies in camel serum and found that the antigen-binding region of the heavy-chain antibody is composed only of a variable region fragment that is functionally equivalent to the antigen-binding fragment (Fab) of conventional antibodies. Therefore, the antigen recognition region fragment of the heavy-chain antibody is called VHH (variable domain of the heavy chain of heavy-chain antibody, VHH), and based on this, a nanobody containing only the VHH domain is developed. Nanobody is also called single-domain antibody (sdAb).

[0037] Nanobodies are easily modified to form multivalent forms. Due to their small molecular weight, nanobodies are encoded by a single gene, are easy to genetically manipulate, and multiple nanobodies can be polymerized through short linker sequences. They can also be linked and combined with conventional antibody Fab fragments, Fv fragments, ScFv fragments, etc. to form multivalent or multispecific antibody structures. Bivalent or multivalent antibodies can recognize the same epitope but have a higher affinity than monovalent antigens. Bispecific or multispecific antibodies can bind to different targets or different binding regions on the same target and have a stronger antigen recognition ability than monovalent antibodies.

[0038] Nanobodies can easily form new fusion molecules with other structures (such as BSA, IgG-Fc, etc.). In the new fusion molecule, the nanobody binds to its target antigen directionally, and the part fused with the nanobody exerts the corresponding function. Therefore, it can be used in combination with other drugs or as a tool for experimental research in diagnosis and various fields. Nanobody screening can be divided into steps such as alpaca immunization, lymphocyte extraction, nanobody library construction, phage library construction, specific phage screening, E. coli expression, and antibody purification.

[0039] The terms "Fc", "Fc segment", or "Fc fragment" refer to the crystallizable fragment that has no antigen-binding activity and is the interaction site between an antibody and an effector molecule or a cell surface Fc receptor (FcR). The Fc fragment contains the constant region polypeptide of an antibody excluding the heavy chain constant region CH1. The Fc fragment binds to cells having the corresponding Fc receptor on their surface and produces various biological effects. In the ADCC effect (antibody-dependent cell-mediated cytotoxicity), the Fab segment of an antibody binds to the antigen epitope of a virus-infected cell or a tumor cell, and its Fc segment binds to the FcR on the surface of killer cells (NK cells, macrophages, etc.), and mediates the killer cells to directly kill the target cell. ADCP is antibody-dependent cellular phagocytosis, and the mechanism of ADCP is that the target cell on which the antibody acts activates the FcγR mechanism on the surface of macrophages, induces phagocytosis, the target cell undergoes internalization, and is acidified and degraded by phagosomes. Under certain circumstances, removal of the antibody Fc function may be beneficial. These circumstances include (1) receptor agonists that induce cell signaling, (2) receptor antagonists that block the binding of a receptor and a ligand to inhibit signaling, or (3) the use of an antibody as a drug vector to deliver a drug to a target cell expressing the corresponding antigen. If the Fc function is maintained, the antibody drug may accidentally damage cells expressing the corresponding receptor, or the antibody-coupled drug may accidentally damage important immune cells when it is off-target.

[0040] Fc variants or combinations of mutations are not limited to the following forms (as determined by EU count):

[0041] [Table 1] TIFF2025518151000003.tif244170TIFF2025518151000004.tif79170

[0042] Currently, mouse-derived antibodies have become a major source of antibody drugs. Since mouse-derived antibodies are immunogenic, they are usually humanized. The following examples provide mouse-derived antibodies, chimeric antibodies, and humanized antibodies. A "chimeric antibody" is an antibody formed by fusing the variable region of a mouse-derived antibody and the constant region of a human antibody, which can reduce the immune response induced by the mouse-derived antibody. The constant region of the human antibody is a heavy chain constant region selected from human IgG1, IgG2, IgG3, IgG4 or variants thereof, and a light chain constant region selected from human κ, λ chains or variants thereof. A "humanized antibody" is an antibody obtained by transplanting the CDR sequences of a mouse-derived antibody into the variable region framework of a human antibody, which can overcome the strong reaction caused by a chimeric antibody that retains a large amount of mouse protein components. Such framework sequences can be obtained from public DNA databases containing germline antibody gene sequences or published reference documents. In order to avoid a decrease in activity due to a decrease in immunogenicity, a minimal number of back mutations or revertant mutations can be performed on the human antibody variable region framework sequence to maintain activity.

[0043] Theoretically, increasing antibody affinity can improve the specificity and efficacy of the antibody, reduce the dosage of the drug, and reduce toxic side effects, etc. In actual research work, it has been proven that there is not necessarily a linear relationship between the increase in affinity and the increase in antibody titer, especially in the treatment of solid tumors. However, in many cases, this linear relationship clearly exists. The humanized antibodies of the present invention also include humanized antibodies that have undergone CDR affinity maturation by phage display. The theoretical basis of in vitro antibody affinity maturation is to mimic the in vivo antibody affinity process. By constructing a random mutation library and mimicking the high-frequency mutations of B cells in the body, high-affinity antibodies can be screened.

[0044] The drug provided by the present invention can contain an antibody or antigen-binding fragment in a "therapeutically effective amount". A "therapeutically effective amount" refers to the amount of a therapeutic agent that effectively prevents or alleviates a specific disease, and can vary based on various factors such as the patient's disease state, age, and weight, as well as the ability of the drug to produce the desired effect in different patients.

[0045] "Sequence identity" refers to the sequence similarity between two polynucleotide sequences or two polypeptides, and is the degree to which two polynucleotides or two polypeptides have the same bases or amino acids. "Having at least 85% sequence identity" as described in the present invention refers to at least 85%, 90%, 95%, 97%, or 99% identity.

[0046] In this specification, amino acid substitutions are named by a one-letter amino acid code, followed by the amino acid position, and then the one-letter amino acid code of the substitution. For example, L234A refers to substituting the L amino acid at position 234 with A.

[0047] The gene encoding SIRPα is a polymorphic gene, and 10 types of SIRPα variants are known in the human population. Katsuto Takenaka et al. sequenced the IgV-encoding SIRPα domain of 37 unrelated normal Caucasians, Africans, Chinese, and Japanese from the Human HapMap Genome Project, and discovered 10 different SIRPα IgV-encoding alleles (Polymorphism in Sirpa modulates engraftment of human hematopoietic stem cells, NATURE IMMUNOLOGY VOLUME 8 NUMBER 12 DECEMBER 2007). The 10 types of SIRPα variants are SIRPα V1 / V2 / V3 / V4 / V5 / V6 / V7 / V8 / V9 / V10 subtypes respectively. SIRPalpha is highly polymorphic, but in the amino acid sequence alignment of known human SIRPalpha alleles by ChiaChiM. Ho et al., there are two unique sequences at the binding interface between SIRPα and CD47, which are allele V1 (a2d1) type and V2 (a1d1) type respectively. (「Velcro」 Engineering of High Affinity CD47 Ectodomain as Signal Regulatory Protein(SIRPα) Antagonists That Enhance Antibody-dependent Cellular Phagocytosis, JOURNAL OF BIOLOGICAL CHEMISTRY, VOLUME 290·NUMBER 20·MAY 15, 2015).

[0048] As shown in FIG. 24, amino acid sequence alignment of known human SIRPα binding domain alleles revealed only two variants, a1d1 and a2d1, at the CD47 contact interface. The first line of text in FIG. 24 is the amino acid sequence of the most important human SIRPα allele V1 (a2d1), and the second line of text in FIG. 24 is the amino acid sequence of the most important human SIRP allele V2 (a1d1). Black boxes indicate residues that interact with CD47, and shading indicates residues that differ from the V1 sequence. Janet Sim et al. used Sanger sequencing of 2535 SIRPα sequences and 510 samples to identify two SIRPα variants, v1 and v2, representing three genetic groups: homozygous v1 / v1, homozygous v2 / v2, and heterozygous v1 / v2. In various populations and unrelated subpopulations, the distribution and frequency of the SIRPα v1 and v2 allele clusters are determined. Here, the distribution of v1 / v2 heterozygosity in five superpopulations of Europe (EUR), the United States (AMR), East Asia (EAS), Africa (AFR), and South Asia (SAS) is similar, with a distribution range of 42.0% - 47.2%. The number of v2 / v2 in the East Asian population is significantly higher than that of v1 / v1, with occurrence frequencies of 42.3% and 13.3% respectively. In Africans, Europeans, Americans, and South Asians, the number of v1 / v1 is higher than that of v2 / v2, and the occurrence frequency ranges of v1 and v2 are 30.3 - 49.1% and 8.9 - 24.2% respectively (for references, see MABS, 2019, VOL.11, NO.6, 1036¨C1052, https: / / doi.org / 10.1080 / 19420862.2019.1624123). Aduro Biotech also studied that the occurrence frequency of v2 / v2 homozygosity in the East Asian population is 41.3%, and the occurrence frequency of v1 / v1 homozygosity is 34.6%. Similarly, it is also proven that 41.3% of the East Asian population is V2 / V2 homozygous (for references, see Voets et al. Journal for ImmunoTherapy of Cancer (2019) 7:340).

[0049] Based on the results of SIRPα polymorphism analysis, the anti-SIRPα antibody can bind to both SIRPα v1 type and SIRPa v2 type genes simultaneously, which is important for promoting clinical development.

[0050] Hereinafter, in combination with specific embodiments, the technical solutions of the present invention will be described in detail. In the following embodiments, experimental methods without specifying specific conditions are based on conventional conditions, or conditions recommended by the manufacturer of raw materials or products, or textbooks of biotechnology such as molecular cloning, laboratory manuals, Cold Spring Harbor Laboratory, modern molecular biology methods, cell biology, etc. Reagents without specific sources indicated are conventional reagents purchased through commercial channels.

[0051] The following Table 1 and Table 2 show the molecules and cell lines used in this study.

[0052]

Table 2

[0053]

Table 3

[0054] Obtaining of anti-SIRPα antibody Example 1: Obtaining of anti-SIRPα mouse antibody (1) Mouse immunization: Anti-human SIRPα monoclonal antibody is generated by immunizing mice. In the experiment, Balb / c white mice, female, 6 weeks old are used. Breeding environment: SPF level. After purchasing the mice, they are bred in the laboratory environment for 1 week, adjusted to a 12 / 12-hour light / dark cycle, with a temperature of 20-25°C and a humidity of 40-60%. The Balb / c mice are immunized. Two weeks after the first immunization of each mouse with 50 μg of recombinant protein QP009 (SIRPα) using complete Freund's adjuvant (CFA), then QP009 (SIRPα) + complete Freund's adjuvant (IFA) or QP009 (SIRPα) + aluminum salt Alum + CpG ODN 1826 are used to immunize each other weekly at 25 μg per mouse.

[0055] QP009 (SIRPα) has the following amino acid sequence (SEQ ID NO: 1).

[0056]

Chemical formula

[0057] (2) Cell fusion: Mice with high antibody titers in the serum are selected for splenocyte fusion. 72 hours before fusion, the selected mice are immunized by intraperitoneal injection of a booster. Using an optimized PEG-mediated fusion step, hybridoma cells are obtained by fusing spleen lymphocytes and myeloma Sp2 / 0 cells. The fused hybridoma cells are resuspended in HAT complete medium (IMDM medium containing 20% FBS, 1×HAT and 1×OPI), dispensed into 96-well cell culture plates (1×105 cells / 150 μl / well), and cultured at 37°C, 5% CO 2 2. On the 5th day after fusion, IMDM medium containing 20% FBS (containing 2×HAT and 1×OPI) is added at 50 μl / well and cultured at 37°C, 5% CO 2 2. On the 7th to 8th day after fusion, according to the cell growth density, all the liquid is replaced, and the medium is HT complete medium (IMDM medium containing 20% FBS, 1×HT and 1×OPI), at 250 μl / well, and cultured at 37°C, 5% CO 2 2.

[0058] (3) Screening of hybridoma cells: According to the cell growth density, on the 10th to 14th day after fusion, ELISA detection is performed to screen for anti-SIRPα antibodies in the hybridoma supernatant. The supernatant is collected from the hybridoma fusion wells, and preliminary screening of the entire 96-well plate is performed by ELISA. The anti-SIRPα antibodies in the detected supernatant can block the binding of SIRPα / CD47, which is the positive well in the preliminary screening. The supernatant is collected from the preliminary screening positive wells, and the binding with QP009 (SIRPα) is detected by ELISA. Clones that are positive for the binding to SIRPα and the blocking of SIRPα / CD47 binding are selected, that is, they are anti-SIRPα antibody positive clone wells. The positive clones are expanded and transferred to a 24 / 6-well plate in a timely manner, and the cell culture supernatant is detected again by ELISA. Clone wells that are positive for the binding to SIRPα and the blocking of SIRPα / CD47 binding are detected, that is, they are anti-SIRPα antibody positive clone wells. The positive clones are serially diluted 2 - 3 times to obtain single cell clones, and the positive single cell strains are cryopreserved to obtain the single cell clone 71C10.

[0059] (4) Sequencing of hybridoma monoclonal antibodies for obtaining antibody sequences: The hybridoma positive monoclonal cell line 71C10 is obtained, mRNA is extracted, mRNA is reverse transcribed into cDNA, the cDNA is used as a template for PCR amplification, PCR positive clones are selected and sent for sequencing, and the variable region sequences of the light and heavy chains of the monoclonal antibody are obtained through sequence analysis. The number and position of CDR amino acid residues conform to the known Kabat numbering rules.

[0060] The variable region sequence of the heavy chain of 71C10 is SEQ ID NO:2, specifically as follows.

[0061]

Chemical formula

[0062] Note: The order is FR1 - CDR1 - FR2 - CDR2 - FR3 - CDR3 - FR4. The bold underlined parts are VHCDR1 (SEQ ID NO:3), VHCDR2 (SEQ ID NO:4), and VHCDR3 (SEQ ID NO:5), respectively.

[0063] The light chain variable region sequence of 71C10 is SEQ ID NO:6, specifically as follows.

[0064]

Chemical formula

[0065] Note: The order is FR1 - CDR1 - FR2 - CDR2 - FR3 - CDR3 - FR4. The bold underlined parts are VLCDR1 (SEQ ID NO:7), VLCDR2 (SEQ ID NO:8), and VLCDR3 (SEQ ID NO:9), respectively.

[0066] Example 2: SPR - detected affinity of anti - SIRPα chimeric antibody (1) Fuse the mouse - derived variable region sequence of monoclonal cell line 71C10 with the human constant region gene to obtain a chimeric antibody molecule. The antibody light chain uses the kappa light chain constant region CL. At the same time, various antigen sequences are designed for the performance test of the antibody molecule. The molecular cloning designs of the antigen and the chimeric antibody are shown in Table 3 and Table 4.

[0067]

Table 4

[0068] Note: Antibodies with protein numbers QP026027 and QP026249 were used as control antibodies, both using the variable region sequences of the known anti-SIRPα antibody KWAR23, and the difference lies in the constant region. QP163164 and QP163245 both use the variable region of the monoclonal cell line 71C10, and the difference lies in the constant region. The sequences shown in the above sequence numbers represent the heavy chain sequence and light chain sequence of each antibody molecule, respectively.

[0069] pQD is the name of a vector combining a signal peptide and a constant region gene (CH1-FC / CL) fragment. Here, pQDH is used for the ligation and expression of the heavy chain variable region, containing a signal peptide and a constant region gene (CH1-FC) fragment. pQDK is used for the ligation and expression of the light chain variable region, containing a signal peptide and a constant region gene (CL) fragment. "H" represents the heavy chain, and "L" represents the light chain. "(IgG4)" indicates that the heavy chain adopts the constant region of human IgG4. When "(IgG4)" is not marked, the constant region of human IgG1 is used by default. 180122VH represents the heavy chain variable region derived from the monoclonal cell line 71C10, and 180122VL represents the light chain variable region derived from the monoclonal cell line 71C10.

[0070] Exemplarily, "pQDH-KWAR23-H" means that the control sequence KWAR23 is fused to the pQDH vector. pQDH contains a signal peptide and a constant region gene (CH1-FC) fragment and uses the constant region of human IgG1. "pQDH-180122VH" means that the heavy chain variable region sequence 180122VH is fused to the pQDH vector and uses the constant region of human IgG1. The sequences shown in the above sequence numbers are specifically as follows.

[0071]

Chemical formula

[0072]

Chemical formula

[0073]

Chemical formula

[0074]

Chemical formula

[0075]

Chemical formula

[0076]

Chemical formula

[0077]

Table 5

[0078] Note: QP098 is the cynomolgus monkey SIRPα sequence (uniprot database accession number I7G9Z7), QP100 is the cynomolgus monkey SIRPα sequence (uniprot database accession number G7PGS8), QP271 is the rhesus monkey SIRPα sequence, obtained by the inventor sequencing cynomolgus monkey PBMC, QP273 is the cynomolgus monkey SIRPα sequence, obtained by the inventor sequencing cynomolgus monkey PBMC.

[0079] (2) Expression and Purification of Antigen and Chimeric Antibody The culture density of 293E cells is maintained at (0.2 - 3)×10 6 / ml, cultured in maintenance medium (GIBCO Freestyle 293 expression medium). One day before transfection, the cells to be transfected are centrifuged, the medium is replaced, and the cell density is adjusted to (0.5 - 0.8)×10 6 / ml. On the day of transfection, the density of 293E cells is (1 - 1.5)×10 6 / ml. Prepare plasmid and transfection reagent PEI. The amount of plasmid required for transfection is 100 μg / 100 ml of cells, and the mass ratio of PEI to plasmid is 2:1. Mix the plasmid and PEI uniformly, let it stand for 15 minutes, and do not exceed 20 minutes. Slowly add the mixture of plasmid and PEI to the 293E cells, place it in a shaker at 8% CO 2 , 120 rpm, 37 °C for culture. On the 5th day after transfection, centrifuge at 4700 rpm for 20 minutes with a horizontal centrifuge, and collect the cell supernatant.

[0080] Protein A affinity chromatography purification: Pass the equilibration solution through the column with at least 3 CV. The actual volume is 20 ml. Confirm that the pH and conductivity of the solution flowing out from the final device are the same as those of the equilibration solution, and the flow rate is 1 ml / min. After centrifugation, pass the culture supernatant through the column, load 40 ml of sample, and the flow rate is 0.33 ml / min. Pass the equilibration solution through the column with at least 3 CV. The actual volume is 20 ml. Confirm that the pH and conductivity of the solution flowing out from the final device are the same as those of the equilibration solution, and the flow rate is 0.33 ml / min. Pass the eluent through the column. When the UV280 rises to 15 mAU, the elution peak (PAC-EP) begins to be collected. When the UV280 drops to 15 mAU, stop the collection, and the flow rate is 1 ml / min. After the sample collection is completed, adjust PAC-EP to neutral with the pH adjustment solution.

[0081] (3) Detection of Affinity by Surface Plasmon Resonance (SPR) The affinity of the anti-SIRPα chimeric antibody QP163164 with human SIRPα V1 type (protein number QP094) and human SIRPα V2 type (protein number QP096) was measured by Biacore T200 (GE). Tables 5 and 6 show the detection results of QP163164 and QP026027. The results show that the SIRPα chimeric antibody QP163164 binds to human SIRPα V1 type with an affinity KD of 5.27E-10 M and binds to human SIRPα V2 type with an affinity KD value of 6.78E-10 M. The binding affinities for human SIRPα V1 type and human SIRPα V2 type are significantly superior compared to the control antibody KWAR23 (QP026027).

[0082]

Table 6

[0083] The affinity of the chimeric antibody and cynomolgus monkey SIRPα measured by biacore is as shown in the following table.

[0084]

Table 7

[0085] Example 3: Humanization of Anti-SIRPα Hybridoma Monoclonal Antibody By aligning the IMGT human antibody heavy and light chain variable region germline gene databases with MOE software, heavy and light chain variable region germline genes with high homology to QP163164 are selected as templates, and the CDRs of mouse-derived antibodies are transplanted into the corresponding human templates respectively to form a variable region sequence in the order of FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4. Several important amino acid residues are selected as combinations of back mutations. Here, the amino acid residues are determined and annotated by the Kabat numbering system. In the following examples, the sequence of the heavy chain FR region is derived from the combined sequence of human germline heavy chain IGHV1-18 and IGHJ2*01, which includes the FR1, FR2, FR3 regions of human germline heavy chain IGHV1-18 and the FR4 region of IGHJ2*01. The sequence of the light chain FR region is derived from the combined sequence of human germline light chain IGKV4-1 and IGKJ2*01, which includes the FR1, FR2, FR3 regions of human germline light chain IGKV4-1 and the FR4 region of IGKJ2*01.

[0086] (1) Humanization molecule cloning of anti-SIRPα antibody Primers for PCR are designed to construct VH / VK gene fragments of each humanized antibody, and homologous recombination is performed with the expression vector pQD (including the signal peptide and constant region gene (CH1-FC / CL) fragment) to construct the full-length antibody expression vector VH-CH1-FC-pQD / VK-CL-pQD.

[0087] Using the online software DNAWorks (v3.2.2) (http: / / helixweb.nih.gov / dnaworks / ), design multiple primers for synthesizing VH / VK containing gene fragments necessary for recombination: 5'-30bp signal peptide + VH / VK + 30bp CH1 / CL-3'. According to the instruction manual of TaKaRa PrimeSTAR GXL DNA Polymerase, use the multiple primers designed above for two-step PCR amplification to obtain VH / VK containing gene fragments necessary for recombination. For the construction and enzymatic cleavage of the expression vector pQD, design and construct the expression vector pQD using some special restriction enzymes such as BsmBI with different recognition sequences and restriction sites. BsmBI is used to enzymatically cleave the vector, cut the gel, recover and store it. Construction of the heavy chain expression vector pQD-VH-CH1-FC and the light chain expression vector pQD-VL-CL: Mix the heavy chain variable region VH gene fragment and the BsmBI enzymatically cleaved vector pQD (containing the signal peptide and the heavy chain constant region (CH1-FC) fragment) at a ratio of 3:1, and mix the light chain variable region VL gene fragment and the BsmBI enzymatically cleaved vector pQD (containing the signal peptide and the light chain constant region (CL) fragment) at a ratio of 3:1. Transfer the mixtures into DH5a competent cells respectively, perform heat shock at 42°C for 90 seconds after incubating in an ice bath at 0°C for 30 minutes, add 5 times the amount of LB medium, incubate at 37°C for 45 minutes, spread on LB-Amp plates, culture overnight at 37°C, select single clones for sequencing to obtain each target clone.

[0088] The following table shows specific information regarding the humanization design of QP163164. The protein expression number is QP256253. In this table, the antibody light chain uses the kappa light chain constant region CL, and the antibody heavy chain uses the human IgG4 constant region (for the specific sequence of the constant region, refer to Example 2). The humanization design of the light chain and heavy chain variable region sequences is not limited to the sequences shown in the following table.

[0089]

Table 8

[0090] Note: The light chain variable region of QP256253 is encoded by the plasmid numbered QD253. The specific sequence of the light chain variable region array SEQ ID NO:16 is as follows.

[0091]

Chemical formula

[0092] The heavy chain variable region of QP256253 is encoded by the plasmid numbered QD256. The specific sequence of the heavy chain variable region array SEQ ID NO:17 is as follows.

[0093]

Chemical formula

[0094] (2) Expression of anti-SIRPα antibody humanized protein The culture density of 293E cells is maintained at (0.2 - 3)×10 6 / ml, and the culture is carried out in maintenance medium (GIBCO Freestyle 293 expression medium). One day before transfection, the cells to be transfected are centrifuged, the medium is exchanged, and the cell density is adjusted to (0.5 - 0.8)×10 6 / ml. On the day of transfection, the density of 293E cells is (1 - 1.5)×10 6 / ml. For the plasmid and the transfection reagent PEI, the amount of plasmid required for transfection is 100 μg / 100 ml of cells, and the mass ratio of PEI to the plasmid is 2:1. The plasmid and PEI are uniformly mixed and left standing for 15 minutes, not exceeding 20 minutes. The mixture of the plasmid and PEI is slowly added to the 293E cells, and 8% CO 2, Incubate in a shaker at 120 rpm and 37 °C. On the 5th day after transfection, centrifuge at 4700 rpm for 20 minutes using a horizontal centrifuge and collect the cell supernatant.

[0095] (3) Purification of anti-SIRPα antibody humanized protein Purification by Protein A affinity chromatography: Pass the equilibration solution through the column with at least 3 CV. The actual volume is 20 ml. Ensure that the pH and conductivity of the solution flowing out from the final device match those of the equilibration solution. The flow rate is 1 ml / min. After centrifugation, pass the culture supernatant through the column, load 40 ml of the sample, and the flow rate is 0.33 ml / min. Pass the equilibration solution through the column with at least 3 CV. The actual volume is 20 ml. Confirm that the pH and conductivity of the solution flowing out from the final device match those of the equilibration solution. The flow rate is 0.33 ml / min. Pass the eluent through the column. When the UV280 rises to 15 mAU, start collecting the elution peak (PAC-EP). When the UV280 drops to 15 mAU, stop collecting. The flow rate is 1 ml / min. After sample collection is complete, adjust the pH of PAC-EP to neutral with a pH adjustment solution.

[0096] (4) Identification of the activity of humanized SIRPα antibody (Binding-ELISA) Binding-ELISA experimental method: Coat QP094 (SIRPαV1-flag-his), QP096 (SIRPαV2-Flag-his), and QP100 (cynoSIRPα-flag-his) at 0.5 μg / ml, 50 μl / well respectively and leave them overnight at 4 °C. Wash 3 times with PBS, incubate with 200 μl / well of 3% BSA / PBS at room temperature for 2 hours, wash 3 times with PBST, add antibodies at different concentrations, incubate at room temperature for 1 hour, wash 3 times with PBST, wash 3 times with PBS, incubate with the secondary antibody HRP-anti Fab diluted 1:2500, incubate at room temperature for 1 hour, wash 3 times with PBST, wash 3 times with PBS, develop with TMB, and stop with 2M 2 SO 4 and read at 450 nm.

[0097] (5) Identification of the Affinity of Humanized SIRPα Antibody by SPR As shown in Table 8 below, the affinity between the humanized antibody and human SIRPα V1 type, human SIRPα V2 type, and cynomolgus monkey SIRPα was measured by biacore. The results show that the anti-SIRPα humanized antibody QP256253 binds to human SIRPα V1 type with an affinity KD of 3.36E-10 M and binds to human SIRPα V2 type with an affinity KD value of 3.19E-10 M.

[0098]

Table 9

[0099] Example 4: Affinity Maturation of Anti-SIRPα Antibody QP163164 (1) Construction of Humanized Phagemid Vector The humanized QP256253 is constructed into a phagemid vector in scFv mode (VH - three GGGGS - VL) respectively as the wild-type sequence (i.e., the original sequence or the starting sequence, the mutant sequence obtained by affinity maturation screening). Over-lap PCR is used to splice VH, (GGGGS)3 linker, and VL, and ligated into the phagemid vector using NcoI and NotI restriction sites.

[0100] (2) Construction of Phage Display Library Using the constructed wild-type scFv as a template, during the primer synthesis process, codon-based primers are used. The codons in the mutant region have 50% wild-type codons and 50% NNK (the reverse primer is MNN), and mutations are introduced into all CDR regions to construct a mutant library. The PCR fragment is enzymatically cleaved with NcoI and NotI, ligated into the phagemid vector, and finally electrotransformed into Escherichia coli TG1. Each codon-based primer independently creates a library.

[0101] (3) Panning of the Library After the library was rescued and the phage particles for panning were packaged, liquid-phase panning was performed using biotinylated QP098 (cynoSIRPα(ECD)) antigen and streptavidin magnetic beads. Also, in each round of screening, the antigen concentration was decreased compared to the previous round. After three rounds of panning, 250 clones were selected for phage ELISA to detect binding activity and sequenced with the positive clones. The sequenced clones were aligned and analyzed, and after removing redundant sequences, the non-redundant sequences were converted into full-length IG (for the heavy-chain constant region, CH1-CH2-CH3 of hIgG4 was selected, and for the light-chain constant region, κ light-chain CL was selected) and expressed in mammalian cells. The full-length IG protein was obtained after affinity purification. The specific sequences are as shown in the following table. In this table, the antibody light chain uses the kappa light-chain constant region CL, and the antibody heavy chain uses the human IgG4 constant region (for the specific sequences of the constant regions, refer to Example 2).

[0102]

Table 10

[0103] Note: The naming rule for protein numbers is the combination of the heavy-chain plasmid number and the light-chain plasmid number. Exemplarily, in the antibody molecule with the protein number QP256279, the heavy-chain plasmid number is QD256, and the light-chain plasmid number is QD279. The sequences shown in the sequence numbers in the table are the sequences of the heavy-chain variable regions or light-chain variable regions of different antibodies. The specific sequences of the light-chain variable regions are as follows.

[0104]

Chemical Structure

[0105]

Chemical Structure

[0106]

Chem.

[0107]

Chem.

[0108]

Chem.

[0109]

Chem.

[0110]

Chem.

[0111]

Chem.

[0112] The bold and underlined parts above are VLCDR1, VLCDR2, and VLCDR3 of each antibody molecule, respectively. The specific comparison with the wild-type sequence QP256253 is as follows.

[0113]

Table 11

[0114] Note: " / " indicates that the sequence is the same as QP256253, and bold and black font indicate amino acids different from QD253. (4) ELISA detection Binding-ELISA experimental method: QP094 (SIRPαV1-flag-his), QP096 (SIRPαV2-Flag-his), QP098 (cynoSIRPα-flag-his), and QP100 (cynoSIRPα-flag-his) at 0.5 μg / ml, 50 μl / well are coated overnight at 4°C. Wash three times with PBS, incubate with 200 μl / well of 3% BSA / PBS for 2 hours at room temperature, wash three times with PBST, add antibodies at different concentrations, incubate for 1 hour at room temperature, wash three times with PBST, wash three times with PBS, incubate with secondary antibody HRP-anti Fab diluted 1:2500, incubate for 1 hour at room temperature, wash three times with PBST, wash three times with PBS, develop with TMB, and stop with 2M H 2 SO 4 and read at 450 nm. The EC50 values are as shown in the following table. The following table further shows the detection results of humanized antibody QP256253, chimeric antibody QP163245, and control antibody QP026249. The results are as shown in Figures 1 to 6.

[0115]

Table 12

[0116] Blocking-ELISA experimental method: Coat QP001 at 2 μg / ml overnight at 4°C, wash three times with PBS, block with 250 μl / well of 5% milk, incubate with a 1:1 mixing ratio of biotin (Biotin)-QP002 at 0.05 μg / ml + Abs at 50 μg / ml for 1 hour at 25°C, and incubate with HRP-streptavidin (1:5000). The results are as shown in Figure 7.

[0117] (5) Detection of affinity by surface plasmon resonance (SPR) The affinity of the anti-SIRPα antibody with human SIRPα V1 type, human SIRPα V2 type, and cynomolgus monkey SIRPα was measured by Biacore, and some of the results are as shown in Table 12. As can be seen from Table 12, the anti-SIRPα antibodies QP2561589, QP2561586, QP2561581, QP256279, and QP2561770 all bind to human SIRPα V1 type and human SIRPα V2 type. At the same time, QP2561589, QP2561586, QP256279, QP2561770, and QP256253 all bind to SIRPα proteins of different cynomolgus monkeys and rhesus monkeys.

[0118]

Table 13

[0119] As can be seen from the above table, the affinity of the affinity-matured antibodies QP2561589, QP2561586, and QP256279 proteins for human SIRPα V1 type and SIRPα V2 type is more than 50 times higher than that of the control antibody KWAR23 (QP026249).

[0120] Example 5: FACS Detection of Anti-SIRPα Antibodies That Bind to Human Renal Clear Cell Carcinoma 786-O Cells That Natively Express Human SIRPα Experimental steps: Collect 2E5 / well of 786-O cells, wash once with PBS, centrifuge at 300 g for 3 minutes, and discard the supernatant. Block: Resuspend in 2% FBS at 2E5 / well, inoculate 200 μl / well into a 96-well U-bottom plate, and hold in an ice bath for 1 hour. Centrifuge at 300 g for 3 minutes and discard the supernatant. Antibody incubation: Incubate with 10 μg / ml of antibody diluted 1:3 at 100 μl / well in an ice bath for 1 hour. Centrifuge and discard the supernatant. Add 200 μl / well of pre-cooled PBS, centrifuge at 300 g for 5 minutes, discard the supernatant, and repeat twice. Secondary antibody: PE-anti-human FC (1:200) at 50 μl / well, in an ice bath for 0.5 hour. Centrifuge and discard the supernatant. Add 200 μl / well of pre-cooled PBS, centrifuge at 300 g for 5 minutes, discard the supernatant, and repeat three times. Read the mean fluorescence value by FACS. The results are as shown in Figure 8, and the SIRPα antibodies QP163245, QP256253, QP256279, QP2561586, and QP2561589 all bind to human renal clear cell carcinoma 786-O cells that naturally express human SIRPα, and the binding affinity is superior to that of the control antibody QP026249 (KWAR23).

[0121] Example 6: In vitro functional experiment of anti-SIRPα antibody ADCP (1) Anti-SIRPα antibodies were prepared into different IgG subtypes, and the molecular cloning design is as follows.

[0122]

Table 14

[0123] Note: The naming rule of protein numbers is a combination of the heavy-chain plasmid number and the light-chain plasmid number. The sequences shown in the heavy-chain accession numbers are the heavy-chain sequences of antibodies of different subtypes. The sequences shown in the light-chain accession numbers are the light-chain or light-chain variable region sequences of antibodies of different subtypes. (L234A, L235A, K338A) indicates that the mutation in the FC segment removes the FCγR function (EU count L234A / L235A / K338A).

[0124] Here, the specific sequence of the heavy chain (SEQ ID NO:26) of QP32700279 is as follows. QVQLVQSGAEVKKPGASVKVSCKASGYTFTNYWINWVRQAPGQGLEWIAMIDPSDSETHYNQIFKDRATLTTDTSTSTAYMELRSLRSDDTAVYYCAMDYGSLYAMDYWGRGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISAAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK。

[0125] The sequence of the variable region of the light chain of QP32700279 is as shown in SEQ ID NO:18. (2) Anti-SIRPα antibody in vitro function experiment ADCP Preparation of macrophages: Resuscitate PBMC, kit EasySep TMMonocytes were isolated using a human monocyte isolation kit (Stemcell-19359), human recombinant M-CSF (final concentration 50 ng / mL) was added, and the mixture was thoroughly and uniformly mixed. The cells were cultured at 37°C for 6 days to induce macrophages, and the cells were collected, counted, and stored. Raji cells were labeled with CFSE. Raji was resuspended to a concentration of 2×10 6 cells / ml, and then 50 μl / well (1×10 5 / well) was added to a 96-well plate containing macrophages. Antibody dilution: Rituximab was diluted to 80 μg / ml in complete medium and serially diluted three-fold to create nine gradients. Anti-SIRPα was diluted to 20 μg / ml in complete medium. Antibody mixing: In the Combination group, the two diluted antibodies were mixed 1:1. In the Rituximab group, an equal volume of medium was mixed. 50 μl / well was added to the 96-well plate of pre-seeded cells and cultured at 37°C for 2 hours. FACS detection: Phagocytosis was measured by gating live CFSE+ / CD14+ cells.

[0126] The affinity matured molecule and control antibody were used in combination with rituximab for the ADCP assay. The experimental results showed that in the combination of SIRPα antibody and rituximab, compared with rituximab alone, the EC50 decreased, indicating an enhanced synergistic effect of ADCP as shown in the original article. The results are as shown in Figures 9, 10, and 11.

[0127] Example 7: Evaluate the inhibitory effect of anti-SIRPα antibody on Raji-Luc tumor growth in the B-NDG-hSIRPΑ mouse model QP32700279 To examine the killing effect of anti-SIRPα antibody against tumors, B-NDG-hSIRPΑ is intravenously inoculated into the Raji-Luc tumor model, and the inhibitory effects of SIRPα antibody and rituximab on tumor growth are evaluated. Raji-Luc cells are cultured in RPMI1640 culture medium containing 10% fetal bovine serum. Raji-Luc cells resuspended in PBS are inoculated into the tail vein of B-NDG-hSIPRa mice at a concentration of 5×10 5 cells / 0.2 mL and a volume of 0.2 mL / mouse. On day 0 and day 3 after inoculation, the tumor imaging signal value is measured using a small animal imager. When the average imaging signal intensity reaches about 1×10 6 P / S, appropriate animal groups are selected according to the tumor imaging signal value and animal body weight, evenly distributed into 4 experimental groups, and each experimental group has 8 animals. Administration is started on the day of grouping, and the specific administration plan is shown in Table 14 below.

[0128]

Table 15

[0129] Note: a: The administration volume is calculated according to 10 μL / g according to the body weight of the experimental animals.

[0130] b: Q3D means administration once every 3 days, and Q2W means administration once every 2 weeks. Counting the day of grouping and administration as D0, at the D18 time point, the tumor growth curve reflected by the tumor imaging signal value of each group and the D18 imaging signal intensity data are as shown in Figure 12 and Table 15.

[0131]

Table 16

[0132] The results of the tumor growth curve show that rituximab, QP32700279, and the combination of QP32700279 and rituximab all significantly inhibited Raji-Luc tumor growth, with tumor inhibition rates (TGI) of 58.6%, 46.4%, and 84.5% respectively, and the combination drug group showing stronger antitumor activity than the single drug group.

[0133] Due to the model characteristics, the mice showed abnormal movements or paralysis in the latter half of the test. At this point, the mice were euthanized and the survival curves were recorded. Until all the mice in group G1 died (D25), the survival curves of each group were as shown in Figure 13.

[0134] For survival analysis, the Kaplan-Meier method was used, and for comparison between groups, the Log rank test was used, with p < 0.05 considered significant. Compared with the control group, both QP32700279 and the combination administration group (QP32700279 + rituximab) could significantly extend the survival period of Raji-Luc tumor-bearing mice (p = 0.0445*, p < 0.001**), while the rituximab group could not effectively extend the survival period of tumor-bearing mice (p = 0.23). The test results suggest that the combined administration of QP32700279 and QP32700279 + rituximab can effectively inhibit the tumor growth of Raji-Luc tumor-bearing mice, thereby improving the survival rate of the mice.

[0135] Example 8: ELISA Detection of Anti-SIRPα Antibodies That Bind to All Subtypes of Human SIRPα According to the SIRPα V1 / V2 / V3 / V4 / V5 / V6 / V7 / V8 / V9 / V10 sequences reported in the literature ("Velcro" Engineering of High Affinity CD47 Ectodomain as Signal Regulatory Protein (SIRPα) Antagonists That Enhance Antibody-dependent Cellular Phagocytosis, JOURNAL OF BIOLOGICAL CHEMISTRY, VOLUME 290·NUMBER 20·MAY 15, 2015), by gene synthesis, the above SIRPα C-terminus was fused to the Fc (mouse IgG2a) of the mouse IgG2a subtype, constructed into the eukaryotic expression vector pQD, and by 293E transient transfection, protein A was used to purify the supernatant on the 5th day of transient transfection, and SIRPα V1 / V2 / V3 / V4 / V5 / V6 / V7 / V8 / V9 / V10 fusion Fc (mouse IgG2a) proteins were obtained respectively. ELISA was further performed to detect the binding of SIRPα antibodies to all subtypes of SIRPα. The sequences are as follows.

[0136] > SIRPα V1(SEQ ID NO:51) GVAGEEELQVIQPDKSVLVAAGETATLRCTATSLIPVGPIQWFRGAGPGRELIYNQKEGHFPRVTTVSDLTKRNNMDFSIRIGNITPADAGTYYCVKFRKGSPDDVEFKSGAGTELSVRAKPS > SIRPαV2(SEQ ID NO:52) GVAGEEELQVIQPDKSVSVAAGESAILHCTVTSLIPVGPIQWFRGAGPARELIYNQKEGHFPRVTTVSESTKRENMDFSISISNITPADAGTYYCVKFRKGSPDTEFKSGAGTELSVRAKPS > SIRPα V3(SEQ ID NO:53) GVAGEEELQVIQPDKSVSVAAGESAILLCTVTSLIPVGPIQWFRGAGPARELIYNQKEGHFPRVTTVSESTKRENMDFSISISNITPADAGTYYCVKFRKGSPD-TEFKSGAGTELSVRAKPS > SIRPα V4(SEQ ID NO:54) GVAGEEGLQVIQPDKSVSVAAGESAILHCTATSLIPVGPIQWFRGAGPGRELIYNQKEGHFPRVTTVSDLTKRNNMDFSIRIGNITPADAGTYYCVKFRKGSPDDVEFKSGAGTELSVRAKPS > SIRPα V5(SEQ ID NO:55) GVAGEEELQVIQPDKFVLVAAGETATLRCTATSLIPVGPIQWFRGAGPGRELIYNQKEGHFPRVTTVSDLTKRNNMDFSIRIGNITPADAGTYYCVKFRKGSPDDVEFKSGAGTELSVRAKPS > SIRPα V6(SEQ ID NO:56) GVAGEEELQVIQPDKSVLVAAGETATLRCTATSLIPVGPIQWFRGAGPGRELIYNQKEGHFPRVTTVSDLTKRNNMDFPIRIGNITPADAGTYYCVKFRKGSPDDVEFKSGAGTELSVRAKPS > SIRPα V7(SEQ ID NO:57) GVAGEEELQVIQPDKSVSVAAGESAILHCTVTSLIPVGPIQWFRGAGPARELIYNQKEGHFPRVTTVSESTKRENMDFSISISNITPADAGTYYCVKFRKGSPD-TEFKSGAGTELSVRGKPS > SIRPα V8(SEQ ID NO:58) GVAGEEELQVIQPDKSVLVAAGETATLRCTATSLIPVGPIQWFRGAGPARELIYNQKEGHFPRVTTVSESTKRENMDFSISISNITPADAGTYYCVKFRKGSPD-TEFKSGAGTELSVRAKPS > SIRPα V9(SEQ ID NO:59) GVAGEEELQVIQPDKSVLVAAGETATLRCTATSLIPVGPIQWFRGAGPGRELIYNQKEGHFPRVTTVSDLTKRNNMDFSIRISNITPADAGTYYCVKFRKGSPDDVEFKSGAGTELSVRAKPS >SIRPα V10(SEQ ID NO:60) RVAGEEELQVIQPDKSVSVAAGESAILHCTVTSLIPVGPIQWFRGAGPARELIYNQKEGHFPRVTTVSESTKRENMDFSISISNITPADAGTYYCVKFRKGSPD-TEFKSGAGTELSVRAKPS >FC (mouse IgG2a)(SEQ ID NO:61) EPRGPTIKPCPPCKCPAPNLLGGPSVFIFPPKIKDVLMISLSPIVTCVVVDVSEDDPDVQISWFVNNVEVHTAQTQTHREDYNSTLRVVSALPIQHQDWMSGKEFKCKVNNKDLPAPIERTISKPKGSVRAPQVYVLPPPEEEMTKKQVTLTCMVTDFMPEDIYVEWTNNGKTELNYKNTEPVLDSDGSYFMYSKLRVEKKNWVERNSYSCSVVHEGLHNHHTTKSFSRTPGK。

[0137] SIRPα antibody to be detected: SIRPα antibody QP256279 was stably expressed in CHO cells, and the number of the CHO stable expression protein is CHO71.

[0138] Construct molecular cloning according to the sequences provided by Patent WO2017178653, and express and purify the anti-SIRPα antibody 18D5 from OSE as an experimental control. At the same time, as described above, the QP026249 is the anti-SIRPα antibody KWAR23 from Forty Seven, which is denoted as KWAR23 herein.

[0139] Experimental steps for ELISA detection of SIRPα antibodies that bind to SIRPα V1 / V2 / V3 / V4 / V5 / V6 / V7 / V8 / V9 / V10: SIRPα V1 / V2 / V3 / V4 / V5 / V6 / V7 / V8 / V9 / V10, 1 μg / ml, 60 μl / well, coat overnight at 4°C, wash twice with PBST, block with 5% fat-free milk (Sangon) 200 μl / well, incubate at room temperature for 1 hour, wash twice with PBST, incubate with antibody 10 μg / ml, 5-fold dilution, 10 gradients, 60 μl / well, incubate at room temperature for 1 hour, wash 5 times with PBST, secondary antibody incubation: anti-hFab1:10000, 60 μl / well, incubate at room temperature for 1 hour, wash 5 times with PBST, color development: Equilibrate TMB at room temperature 1 hour before, develop color for 10 minutes at 100 μl / well, 2M H 2 SO 4 Stop at 50 ul / well and read at 450 nm with a microplate reader.

[0140] The experimental results are as shown in Figures 14 to 23. The SIRPα antibody CHO71 of the present invention binds to all subtypes of SIRPα V1 / V2 / V3 / V4 / V5 / V6 / V7 / V8 / V9 / V10. The SIRPα antibody 18D5 from OSE does not bind to SIRPα V2 / V3 / V7 / V8 / V10.

[0141] Construction and detection of bispecific antibodies targeting SIRPα and PD-L1 Based on the above results, a bispecific antibody targeting SIRPα and PD-L1 is constructed using the sequence of the anti-SIRPα antibody with protein number QP256279. The SIRPα-binding domain of the bispecific antibody includes a heavy chain variable region and a light chain variable region. The sequence of the heavy chain variable region is selected from QD256, and the sequence of the light chain variable region is selected from QD279. By combining with the sequence of the PD-L1 nanobody initially obtained by the applicant (invention name: anti-PD-L1 nanobody and its use, patent publication number: CN112574309A, application number: 202011309419.7), a bispecific antibody targeting SIRPα and PD-L1 is constructed. The plasmid number of the selected PD-L1 nanobody is QD509, which is obtained by immunizing and humanizing alpacas. In this study, QD509 can represent the VHH fragment of anti-PD-L1 and can also represent the fusion protein of the VHH fragment and FC. The amino acid sequence of the VHH fragment is as shown in SEQ ID NO:62, specifically as follows.

[0142]

Chemical Structure

[0143] Example 9: Clone Design An anti-SIRPα / PD-L1 bispecific antibody molecule was designed, and the anti-PD-L1 nanobody QP509 VHH was fused to the C-terminus of the SIRPα antibody heavy chain (e.g., QD3282 in Table 16), or the N-terminus (e.g., QD626 in Table 16), via G4S linker sequences with different repeat numbers, or fused to the N-terminus of the SIRPα antibody light chain (e.g., QD623 in Table 16). Primers were designed according to the sequences, and the full length of each bispecific antibody gene designed by PCR was constructed. Homologous recombination was performed with the expression vector pQD, and each constructed expression vector pQD was numbered with a plasmid number as shown in Table 16, and the obtained antibody molecules were numbered with a protein number. In Table 16, the amino acid sequences of QD623, QD624, and QD625 are basically similar, and the difference lies in the number of G4S repeats. Similarly, the amino acid sequences of QD626, QD627, and QD628 are basically similar, and the difference lies in the number of G4S repeats. The bispecific antibody sequences and protein expression numbers are as follows.

[0144]

Table 17

[0145] In Table 16, the specific sequences of SEQ ID NO:66 and SEQ ID NO:67 that make up QP32820279 are as follows.

[0146] >QD3282(SEQIDNO:66) QVQLVQSGAEVKKPGASVKVSCKASGYTFTNYWINWVRQAPGQGLEWIAMIDPSDSETHYNQIFKDRATLTTDTSTSTAYMELRSLRSDDTAVYYCAMDYGSLYAMDYWGRGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISAAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGGGGGSGGGGSGGGGSGGGGSEVQLLESGGGLVQPGGSLRLSCAASGFTLDDYAIGWFRQAPGKGREGVSCISKSGETTFFVESVKDRFTISRDNSKNTLYLQMNSLRAEDTAVYYCATGSWCTVGSMSRQFYRQFFHSWGQGTLVTVSS*。

[0147] >QD279(SEQIDNO:67) DIVLTQSPDSLAVSLGERATINCRASQSVRSSGYNWIFWYQQKPGQPPKLLIYLASNRDSGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQHSRELPTFGQGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC。

[0148] The monoclonal antibody against SIRPα and the anti-PD-L1 nanobody are designed to fuse with the FC molecule and are used as controls. The protein number of the monoclonal antibody against SIRPα is QP32700279, and the FC molecule numbers of the anti-PD-L1 nanobody fusion are QP509 and QP3447, specifically as shown in Table 17.

[0149]

Table 18

[0150] Furthermore, based on existing literature, the anti-SIRPα antibody and anti-PD-L1 antibody used as controls are further designed. Specifically, it is as shown in Table 18. Here, QP026249 is an analog of the SIRPα monoclonal antibody KWAR23 from 47 companies. QP250251 is an analog of the anti-SIRPα monoclonal antibody 18D5 from OSE Immunotherapeutics. QP37503751 is an analog of the anti-SIRPα monoclonal antibody 1H9 from 47 companies. QP11801181 is an analog similar to Tencentriq, and Tencentriq is the anti-PD-L1 monoclonal antibody Atezolizumab from Roche.

[0151]

Table 19

[0152] Based on existing literature, the SIRPα antigen is designed for experiments, as shown in Table 19.

[0153]

Table 20

[0154] Note: QP098 is the cynomolgus monkey SIRPα sequence (Uniprot database accession number I7G9Z7), QP271 is the rhesus monkey SIRPα sequence, and was obtained by the inventor sequencing cynomolgus monkey PBMC. QP532 - QP538 represent sequences encoding the SIRPα molecule, and the mFC sequence is as shown in SEQ ID NO:86.

[0155] Example 10: Protein Expression and Purification Referring to the protein expression and purification method of Example 2, the protein of Example 9 is expressed and purified. Each antibody is further purified by SEC, and the results show that the transient expression yield of the anti - SIRPα / PD - L1 bispecific antibody is good, the SEC purity is good, and the physical and chemical properties are stable.

[0156] Example 11: SPR Detection of Anti - SIRPα / Anti - PD - L1 Bispecific Antibody Binding to SIRPα / PD - L1 Affinity (1) In this study, the affinity between the antibody molecule and the antigen SIRPα was detected through Biacore8K. As described above, QP026249 is an analog of the SIRPα monoclonal antibody KWAR23 of 47 companies. The results are shown in the following table.

[0157] [Table 21]

[0158] The results show that all different forms of bispecific antibody molecules designed in the present invention bind to the SIRPα V1 recombinant protein with high affinity. (2) Next, the affinity between the antibody molecule and the antigen PD - L1 was detected through Biacore8K. QP11801181 is an analog of the Roche PD - L1 monoclonal antibody Tecentriq. QP509 is the N - terminus where the PD - L1 nanobody VHH is fused to FC, and QP3447 is the C - terminus where the PD - L1 nanobody is fused to FC. The results are shown in the following table.

[0159]

Table 22

[0160] The results show that different forms of bispecific antibody molecules designed in this invention bind to human PD-L1 recombinant protein with high affinity, where the C-terminal affinity of PD-L1 nanobody VHH fused to FC is slightly higher than that in the form fused to the N-terminal.

[0161] Example 12: ELISA detection of anti-SIRPα / anti-PD-L1 bispecific antibody that binds to PD-L1 and SIRPα proteins (1) ELISA detection of the binding between the antibody and PD-L1: Coating plate: Anti-his, 1 μg / ml in PBS, 60 μl / well, overnight at 4°C, wash twice with PBST. Block: 5% non-fat milk (non-fat milk, Sangon), 200 μl / well, 25°C, 120 rpm, incubate for 1 hour. Antigen QP003 (PDL1-his), 1 μg / ml, 60 μl / well, 25°C, 120 rpm, incubate for 1 hour, wash 5 times with PBST. Dilute the primary antibody 5-fold starting from 13.3 nM, seven gradients, and the last 10-fold dilution, 60 μl / well, 25°C, 120 rpm, incubate for 1 hour, wash 5 times with PBST. Secondary antibody anti-hFc1:5000, 60 μl / well, 25°C, 120 rpm, incubate for 1 hour, wash 5 times with PBST. Color development: Equilibrate TMB at room temperature 1 hour in advance, develop color for 3 minutes, Stop: 1 M H 2 SO 4 to stop the color development. The results are as shown in Figure 25.

[0162] (2) ELISA detection of the antibody molecule that binds to SIRPα: Coating plates: Coat each of QP093 (SIRPαV1) and QP095 (SIRPaV2) overnight at 4°C with 1 μg / ml in PBS, 60 μl / well, and wash twice with PBST. Block: Incubate with 5% non-fat milk (Sangon), 200 μl / well, at 25°C, 120 rpm for 1 hour. Dilute the primary antibody 5-fold at 66.7 nM, in seven gradients, with the last dilution at 50-fold, 60 μl / well, incubate at 25°C, 120 rpm for 1 hour, wash 5 times with PBST. Incubation with secondary antibody: Anti-hFab (without glycerol) 1:10000, 60 μl / well, incubate at 25°C, 120 rpm for 1 hour, wash 5 times with PBST. Color development: Equilibrate TMB to room temperature 1 hour prior, develop color for 3 minutes. Stop: Stop color development with 1 M H 2 SO 4 to stop color development. The results are as shown in Figures 26 and 27.

[0163] Example 13: ELISA Detection of Antibody Molecules that Block the Binding between Human PD-L1 and PD-1 Protein Coat the protein QP1138 (PD1-FC) at 2 μg / ml, 50 μl / well, overnight at 4°C. Wash 3 times with PBS. Block: Incubate with 3% BSA, 250 μl / well, at room temperature for 1 hour. Prepare 2 μg / ml of PDL1-mouse FC and different concentrations of antibodies respectively, mix uniformly in equal volume, and incubate at room temperature for 1 hour. Wash 3 times with PBST and 3 times with PBS. Incubation with secondary antibody: HRP-mouse IgG (1:5000), 50 μl / well, wash 6 times with PBST and 3 times with PBS. Color development: Develop color with TMB at 100 μl / well for 10 minutes. Stop with 2 M H 2 SO 4 to stop at 50 μl / well. The results are as shown in Figure 28, and the dual antibody molecule can block the binding between human PD-L1 and PD-1 protein.

[0164] Example 14: ELISA Detection of Antibody Molecules that Block the Binding between Human CD47 and SIRPα Protein Blocking-ELISA experimental method: Coat with 2 μg / ml of CD47-FC (QP001) overnight at 4°C, wash three times with PBS, block with 250 μl / well of 5% milk, incubate with a 1:1 mixture of 0.05 μg / ml of biotin-SIRPα-FC (QP002) + 50 μg / ml of Abs, and incubate with HRP-streptavidin (1:5000) at 25°C for 1 hour. The results are as shown in Figure 29.

[0165] Example 15: Synergistic enhancement of the CD20 antibody rituximab-dependent ADCP effect Antibody-dependent macrophage-mediated phagocytosis (ADCP) refers to the binding of the Fab segment of an antibody to the antigen epitope of a tumor cell, and its Fc segment binds to the FcγR on the surface of macrophages, mediating macrophages to phagocytose target cells. At the same time, SIRPα expressed on macrophages binds to CD47 expressed on tumor cells to form an inhibitory signal. Through ADCP, the bispecific antibody molecule blocks the binding of macrophage SIRPα and Raji cell CD47, thereby further studying the synergistic enhancement of the biological activity of CD20 antibody rituximab-dependent macrophages on human Burkitt lymphoma cells Raji.

[0166] Isolate monocytes from the peripheral blood mononuclear cells (PBMC) of healthy individuals, add 50 ng / mL of human recombinant M-CSF to induce macrophage differentiation. Label Raji cells with green fluorescent CFSE, inoculate Raji cells and macrophages in a 96-well plate at a ratio of 2:1, and add various concentrations of CD20 antibody rituximab alone or in combination with the SIRPα antibody molecule. After incubating at 37°C for 2 hours, stop the reaction, incubate with APC anti-human CD11b antibody, and read through FACS. The percentage of APC / FITC double-positive cells at each concentration of the antibody is the percentage of macrophages undergoing phagocytosis.

[0167] The results are as shown in Figure 30. The results indicate that the combined use of the antibody of the present invention and rituximab is smaller than the single EC50 with rituximab, and the ADCP synergistic effect is significantly enhanced.

[0168] Example 16: Biological Activity of Bispecific Antibody Molecules that Stimulate Human PBMC Proliferation in Vitro Human peripheral blood mononuclear cells (PBMC) are composed of various white blood cells mainly including monocytes, B cells, T cells, NK cells, dendritic cells, macrophages, etc. PBMC are stimulated by adding superantigen SEB in vitro, and lymphocytes are activated and proliferated through the presentation and activation of APC cells among them, generating various cytokines. The PD-L1 antibody promotes the proliferation of T cells and the release of cytokines such as IL-2 by blocking the immune inhibitory signal of PD-1 / PD-L1 binding. The amount of IL-2 released is detected through ELISA to further study the biological activity of the antibody molecule in the in vitro proliferation experiment of PBMC. PBMC cells are inoculated into 96-well plates, different concentrations of SEB are prepared and added to the wells of PBMC cells, then the antibody molecule of the present invention and other control antibodies are added, gently mixed uniformly, and cultured for 2 days. The secretion amount of IL-2 in the cell culture supernatant is detected by ELISA. The results show that QP32700624 and QP32820279 can significantly enhance the activation and proliferation of PBMC and enhance the production of IL-2 in the in vitro proliferation experiment of SEB-stimulated PBMC, comparable to the control antibody tencentriq. The results are as shown in Figure 31.

[0169] Example 17: Production and Purification of Q-1801 Protein The sequence of the anti-SIRPα / anti-PD-L1 bispecific antibody QP32820279 was transferred into a pCHO vector, which is a laboratory-modified vector containing GS as a screening marker and can be used for screening stably transfected CHO cells. CHO cells were stably transfected and pressured and screened by GS to obtain a cell line highly expressing the anti-SIRPα / anti-PD-L1 bispecific antibody for protein production, which was purified to obtain the target molecule, numbered CHO44 (named Q-1801), and the sequence number of the CHO44 protein is the same as that of QP32820279 (the amino acid sequence numbers are SEQ ID NO: 66 and SEQ ID NO: 67). At the same time, the SIRPα monoclonal antibody QP32700279 was also expressed and purified in CHO cells, numbered CHO71, and used as a single-component control for SIRPα. The PD-L1 nanobody VHH was fused to the C-terminus of FC, with the protein number QP3447, and used as a single-component control for PD-L1.

[0170] Example 18: Q-1801 that Binds to Human and Monkey SIRPα (1) SPR The gene encoding human SIRPα is a polymorphic gene, and 10 variants have been identified in the human population. The polymorphism of human SIRPα causes changes in surface-exposed amino acids but does not affect binding to CD47. The most common protein variants are SIRPαV1 and V2 (accession numbers NP_542970 (P78324) and CAA71403). In this study, Biacore8K was used to detect the affinity between molecules such as Q-1801 and the antigen SIRPα. The results of the binding affinities of Q-1801, QP026249 with human SIRPαV1, and Q-1801 with human SIRPαV2, cynomolgus monkey SIRPα, and rhesus monkey SIRPα are shown in Table 22. The results show that Q-1801 binds to human SIRPαV1 with high affinity, with a KD value of 6.01E-11 (M), the KD value of binding to SIRPαV2 is 1.03E-10 (M), the KD value of binding to cynomolgus monkey SIRPα is 2.14E-09 (M), and the KD value of binding to rhesus monkey SIRPα is 5.22E-10 (M). The affinity of Q-1801 for binding to human SIRPαV1 is significantly higher than that of the KWAR23 analog (QP026249). The KD value of the affinity of the KWAR23 analog for binding to human SIRPαV1 is 4.69E-09 M. The affinity of Q-1801 for binding to human SIRPαV2 is significantly higher than that of the KWAR23 analog (QP026249). The KD value of the affinity of the KWAR23 analog for binding to human SIRPαV2 is 1.75E-08 (M).

[0171]

Table 23

[0172] (2) ELISA Based on the SIRPα gene polymorphisms in different ethnic groups and existing literature reports, the ratios of SIRPαV1 / V2 are different in each ethnic group. Here, the SIRPαV2 gene in East Asians reaches 42.3%. ELISA was used to detect the binding of molecules such as Q-1801 to different subtypes of human SIRPα and mouse SIRPα.

[0173] The inventors constructed nine previously reported SIRPα genotypes V1 to V9. Human SIRPα V1 recombinant protein (QP093), human SIRPα V2 recombinant protein (QP095), human SIRPα V3 recombinant protein (QP532), human SIRPα V4 recombinant protein (QP533), human SIRPα V5 recombinant protein (QP534), human SIRPα V6 recombinant protein (QP535), human SIRPα V7 recombinant protein (QP536), human SIRPα V8 recombinant protein (QP537), and human SIRPα V9 recombinant protein (QP538) were coated on ELISA plates, and test molecules such as Q-1801 at gradient dilutions were added, detected with HRP-labeled anti-human Fc secondary antibody, and the results were as shown in FIGS. 32 to 40. Q-1801 bound to human SIRPα V1 with an EC50 = 0.1768 nM, bound to human SIRPα V2 with an EC50 = 0.2101 nM, bound to human SIRPα V3 with an EC50 = 0.1543 nM, bound to human SIRPα V4 with an EC50 = 0.1631 nM, bound to human SIRPα V5 with an EC50 = 0.1667 nM, bound to human SIRPα V6 with an EC50 = 0.2721 nM, bound to human SIRPα V7 with an EC50 = 0.2182 nM, bound to human SIRPα V8 with an EC50 = 0.4176 nM, and bound to human SIRPα V9 with an EC50 = 0.3991 nM. The 18D5 analog did not bind to human SIRPα V2, did not bind to human SIRPα V3, did not bind to human SIRPα V7, and did not bind to human SIRPα V8.

[0174] In summary, Q-1801 has high affinity for all genotypes of SIRPα, while QP250251 (18D5 analog) does not bind to SIRPα V2 / V3 / V7 / V8 / V10, and QP026249 (KWAR23 analog) has a weaker binding to SIRPα V2 / V3 / V7 / V8 / V10.

[0175] (3) FACS According to existing literature reports, the U-937 cell line is a human histiocytic lymphoma cell that expresses endogenous SIRPαV1, and the THP-1 cell line is a human monocytic leukemia cell that expresses endogenous SIRPαV2. By FACS measurement, the binding of Q-1801 to human SIRPαV1 and human SIRPαV2 proteins is tested. Molecules such as Q-1801 at different concentrations are incubated with U-937 cells and THP-1 cells respectively, the fluorescence values are detected by FACS, a fitting curve is plotted, and the EC50s are compared.

[0176] The results are as shown in Figure 41. The EC50 of Q-1801 binding to U-937 (human SIRPαV1) is 0.1057 nM, the CHO71 binding EC50 is 0.07598 nM, the QP026249 (KWAR23 analog) binding EC50 is 0.21 nM, the QP37503751 (1H9 analog) binding EC50 is 0.1583 nM, and the QP250251 (18D5 analog) binding EC50 is 0.5811 nM. The affinity of Q-1801 binding to human SIRPαV1 is comparable to that of the monoclonal antibody CHO71.

[0177] The results are as shown in Figure 42. The EC50 of Q-1801 binding to THP-1 (human SIRPαV2) is 0.1037 nM, the CHO71 binding EC50 is 0.0743 nM, the QP026249 (KWAR23 analog) binding EC50 is 0.2606 nM, the QP37503751 (1H9 analog) binding EC50 is 0.2103 nM, and it does not bind to QP250251 (18D5 analog). The affinity of Q-1801 binding to human SIRPαV2 is comparable to that of the monoclonal antibody CHO71.

[0178] In summary, Q-1801 binds to human SIRPαV1 and human SIRPαV2 with high affinity. QP250251 (18D5 analog) does not bind to human SIRPαV2.

[0179] Example 19: Q-1801 That Binds to Human PD-L1 and Monkey PD-L1 (1) SPR The affinity between molecules such as Q-1801 and the antigen PD-L1 is detected by Biacore8K. The affinity results of Q-1801, QP3447, QP11801181 (a tesentriq analog) and human PD-L1, and the affinity result of Q-1801 and cynomolgus monkey PD-L1 are shown in Table 23.

[0180] [Table 24]

[0181] The results show that the KD value of Q-1801 binding to human PD-L1 with high affinity is 4.31E-10 (M), and the KD value of binding to cynomolgus monkey PD-L1 is 4.99E-10 (M). The KD value of QP3447 binding to human PD-L1 is 4.79E-10 (M). The KD value of the tesentriq analog binding to human PD-L1 is 1.39E-09 (M).

[0182] In summary, both Q-1801 and QP3447 bind to human PD-L1 with high affinity, and their affinities are equivalent. The affinity of Q-1801 binding to human PD-L1 is higher than that of the PD-L1 positive antibody tesentriq analog.

[0183] (2) ELISA The binding between molecules such as Q-1801 and human PD-L1 is detected by ELISA.

[0184] The anti-HIS antibody is coated on the ELISA plate, different species of PD-L1 proteins are added, and after incubation, molecules such as serially diluted Q-1801 are added. The HRP-labeled anti-human Fc secondary antibody is detected, and the results are as shown in Figure 43. Q-1801 binds to the human PD-L1 protein, and the EC50 value is 0.1218 nM. QP3447 binds to the human PD-L1 protein, and the EC50 value is 0.08847 nM. Tesentriq binds to the human PD-L1 protein, and the EC50 value is 0.09194 nM.

[0185] (3)FACS According to existing literature reports, the HCC827 cell line is a human lung cancer cell that expresses endogenous PD-L1. By FACS, the binding of Q-1801 to HCC827 cells that naturally express human PD-L1 protein is tested. Molecules such as Q-1801 at different concentrations and HCC827 cells are incubated, the fluorescence value is detected by FACS, a fitting curve is plotted, and the EC50 is compared. The results are as shown in Figure 44. The EC50 for Q-1801 to bind to human PD-L1 is 0.1416 nM, the QP3447 binding EC50 is 0.1188 nM, and the tislelizumab binding EC50 is 0.1089 nM.

[0186] In summary, Q-1801, the PD-L1 monoclonal antibody QP3447, and tislelizumab all bind to HCC827 cells that naturally express human PD-L1, and the binding affinities are equivalent.

[0187] Example 20: Q-1801 that blocks the binding of SIRPα and CD47 The CD47 protein overexpressed on the surface of tumor cells binds to SIRPα expressed on the surface of macrophages, evading phagocytosis by macrophages. The Q-1801 molecule may block the binding between CD47 and SIRPα, causing the loss of the "don't eat me" signal and promoting macrophage tumor attack. Competitive ELISA is used to detect the ability of molecules such as Q-1801 to block the binding between SIRPα and CD47. Human CD47 protein is coated on an ELISA plate, SIRPα-mouse Fc protein is added, and after incubation, serially diluted molecules such as Q-1801 are added, and an HRP-labeled anti-mouse IgG secondary antibody is detected. The results are as shown in Figure 45. The IC50 of Q-1801 blocking the binding between SIRPα and CD47 is 0.8149 nM, the IC50 of CHO71 blocking the binding between SIRPα and CD47 is 0.7074 nM, the IC50 of QP026249 (KWAR23 analog) blocking the binding between SIRPα and CD47 is 3.12 nM, the IC50 of QP37503751 (1H9 analog) blocking the binding between SIRPα and CD47 is 2.277 nM, and the IC50 of QP250251 (18D5 analog) blocking the binding between SIRPα and CD47 is 32.98 nM.

[0188] In summary, the ability of Q-1801 to block the binding between SIRPα and CD47 is comparable to that of the SIRPα monoclonal antibody CHO71 and is significantly more potent than the abilities of QP026249 (KWAR23 analog), QP37503751 (1H9 analog), and QP250251 (18D5 analog).

[0189] Example 21: Q-1801 Blocking PD-L1 / PD-1 and PD-L1 / CD80 Binding PD-L1 has two ligands, PD-1 and CD80. The C-terminus of the Q-1801 molecule is an anti-PD-L1 nanobody that not only blocks the binding between PD-L1 and PD-1 but also blocks the binding between CD80 and PD-L1.

[0190] By competitive ELISA, molecules such as Q-1801 that block the binding of PD-1 and PD-L1 are detected. Human PD-1 protein is coated on an ELISA plate, PD-L1-mouse Fc protein is added, and after incubation, serially diluted molecules such as Q-1801 are added, and an HRP-labeled anti-mouse IgG secondary antibody is detected. The results are as shown in Figure 46. The IC50 of Q-1801 that blocks PD-L1 and PD-1 is 0.9043 nM, the IC50 of QP3447 that blocks PD-L1 and PD-1 is 0.9511 nM, and the IC50 of tesentriq that blocks PD-L1 and PD-1 is 2.422 nM.

[0191] By competitive ELISA, the ability of molecules such as Q-1801 that block the binding of CD80 and PD-L1 is detected. Human CD80 protein is coated on an ELISA plate, PD-L1-mouse Fc protein is added, and after incubation, serially diluted test molecules such as Q-1801 are added, and an HRP-labeled anti-mouse IgG secondary antibody is detected. The results are as shown in Figure 47. The IC50 of Q-1801 that blocks PD-L1 and CD80 is 0.7415 nM, the IC50 of QP3447 that blocks PD-L1 and CD80 is 0.746 nM, and the IC50 of tesentriq that blocks PD-L1 and CD80 is 1.683 nM.

[0192] In summary, Q-1801 not only blocks the binding of PD-L1 / PD-1 but also blocks the binding of PD-L1 / CD80, and the ability of Q-1801 to block the binding of PD-L1 / PD-1 and PD-L1 / CD80 is superior to that of tesentriq.

[0193] Example 22: Q-1801 that synergistically enhances the CD20 antibody rituximab-dependent ADCP effect Antibody-dependent macrophage-mediated phagocytosis (ADCP) refers to the Fab segment of an antibody that binds to the antigen epitope of tumor cells, and its Fc segment binds to FcγR on the surface of macrophages, mediating the phagocytosis of target cells by macrophages. However, SIRPα expressed on macrophages binds to CD47 expressed on tumor cells to form an inhibitory signal. Raji cells are human Burkitt lymphoma cells that endogenously express CD47 / CD20. Further research is conducted to investigate whether ADCP blocks the binding of Q-1801 to macrophage SIRPα and Raji cell CD47, thereby synergistically enhancing the biological activity of CD20 antibody rituximab-dependent macrophages against human Burkitt lymphoma cells Raji.

[0194] Monocytes are isolated from peripheral blood mononuclear cells (PBMCs) from two different donors and induced to differentiate into macrophages by adding 50 ng / mL of human recombinant M-CSF. Raji cells are labeled with green fluorescent CFSE, and Raji cells and macrophages are seeded in a 96-well plate at a ratio of 2:1, and different concentrations of the CD20 antibody rituximab are added alone or in combination with an SIRPα antibody molecule such as Q-1801. After incubating at 37 °C for 2 hours, the reaction is stopped, and APC anti-human CD11b antibody is incubated. Through FACS reading, the percentage of APC / FITC double-positive cells at each concentration of the antibody is obtained, which is the percentage of macrophages that generate phagocytosis.

[0195] The results are as shown in Figure 48. Donor: P121031405C, Rituxan caused macrophages to phagocytize Raji cells in a concentration-dependent manner. The maximum phagocytosis rate was approximately 26.57%, the EC50 value was approximately 0.02115 μg / mL. The effect of Q-1801 synergistically acting with Rituxan-dependent macrophages to phagocytize human Burkitt lymphoma Raji cells was stronger. The maximum phagocytosis rate increased from 26.57% to 32.38%, and the EC50 value was approximately 0.01188 μg / mL. The single-component control CHO71 was comparable to Q-1801. The maximum phagocytosis rate increased from 26.57% to 32.14%, and the EC50 value was approximately 0.01302 μg / mL. The maximum phagocytosis rate of the positive control QP026249 (KWAR23 analog) increased from 26.57% to 30.09%, and the EC50 value was approximately 0.01765 μg / mL. The maximum phagocytosis rate of the positive control QP37503751 (1H9 analog) was approximately 28.07%, and the EC50 value was approximately 0.01485 μg / mL.

[0196] The results are as shown in Figure 49. Donor: P121070501C, Rituxan caused macrophages to phagocytize Raji cells in a concentration-dependent manner. The maximum phagocytosis rate was approximately 31.32%, the EC50 value was approximately 0.0578 μg / mL. The combined result of Q-1801 + Rituxan was that Q-1801 significantly increased the phagocytosis of Rituxan-dependent macrophages of Raji cells. The maximum phagocytosis rate was approximately 44.17%, and the EC50 value was approximately 0.02733 μg / mL. The maximum phagocytosis rate of the single-component control CHO71 was approximately 43.66%, and the EC50 value was approximately 0.02784 μg / mL. The maximum phagocytosis rate of the positive control QP026249 (KWAR23 analog) was approximately 40.71%, and the EC50 value was approximately 0.04938 μg / mL. The maximum phagocytosis rate of the positive control QP37503751 (1H9 analog) was approximately 37.73%, and the EC50 value was approximately 0.03626 μg / mL.

[0197] In summary, the combination of Q-1801 and Rituxan significantly increases the phagocytosis of Rituxan-dependent macrophages of Raji cells. Example 23: Q-1801 that can stimulate the proliferation of T cells in the mixed lymphocyte reaction The mixed lymphocyte reaction refers to the co-culture of human T cells and allogeneic dendritic cells. Lymphocytes are activated and proliferated by stimulation from allogeneic antigens, generating a variety of cytokines. The PD-L1 antibody blocks the immune inhibitory signal of PD-1 / PD-L1 binding in an antibody concentration-dependent manner, stimulates T cell proliferation, and releases cytokines such as IL-2 / IFN-γ. Detect the release amount of IL-2 / IFN-γ by ELISA and further study the biological activity of Q-1801 in stimulating in vitro T cell proliferation in the mixed lymphocyte reaction.

[0198] Isolate monocytes in PBMC, add rhGM-CSF and rhIL-4 to induce DC (induced dendritic cells), and isolate CD4+ T cells in PBMC from another donor. Mix DC cells and T cells at a ratio of 1:10, add antibodies at different concentrations, perform mixed culture for 2 - 5 days, and detect the expression of IL-2 and IFN-γ in the culture supernatant. The results show that Q-1801, the single-component control QP3447, and the control antibody Tencentriq can all stimulate T cell proliferation and enhance the production of IL-2 and IFN-γ in the mixed lymphocyte reaction (MLR). (See Figures 50 and 51).

[0199] In summary, Q-1801 can stimulate T cell proliferation and enhance the production of IL-2 and IFN-γ in the mixed lymphocyte reaction (MLR), and the IFN-γ secretion amount is superior to that of Tencentriq.

[0200] Example 24: Q-1801 that stimulates the biological activity of human PBMC in vitro proliferation Human peripheral blood mononuclear cells (PBMCs) are mainly composed of various white blood cells including monocytes, B cells, T cells, NK cells, dendritic cells, macrophages, etc. In vitro, PBMCs are stimulated by adding superantigen SEB, and lymphocytes are activated and proliferated through the presentation and activation of APC cells among them, generating a variety of cytokines. The PD-L1 antibody enhances the proliferation of T cells and the release of cytokines such as IL-2 / IFN-γ by blocking the immune inhibitory signal of PD-1 / PD-L1 binding. The release amount of IL-2 / IFN-γ is detected by ELISA to further study the biological activity of Q-1801 in the in vitro proliferation experiment of PBMCs. PBMC cells are inoculated into 96-well plates, different concentrations of SEB are prepared, added to PBMC cell wells, Q-1801 and other control antibodies are added, gently and uniformly mixed, and cultured for 2 - 5 days. The secretion amount of IL-2 in the cell culture supernatant is detected by ELISA, and the secretion amount of IFN-γ in the cell culture supernatant is detected by ELISA. The results show that in the in vitro proliferation experiment of SEB-stimulated PBMCs, Q-1801 can significantly enhance the activation and proliferation of PBMCs and enhance the production of IL-2 and IFN-γ. The single-component control QP3447 can significantly enhance the activation and proliferation of PBMCs and enhance the production of IL-2 and IFN-γ in the in vitro proliferation experiment of SEB-stimulated PBMCs. The control antibody Tencentriq can significantly enhance the activation and proliferation of PBMCs and enhance the production of IL-2 and IFN-γ in the in vitro proliferation experiment of SEB-stimulated PBMCs. The results are as shown in Figures 52 and 53.

[0201] In summary, in the in vitro proliferation experiment of SEB-stimulated PBMCs, Q-1801 can significantly enhance the activation and proliferation of T cells and enhance the production of IL-2 and IFN-γ, and its activity is comparable to that of Tencentriq.

[0202] Example 25: Inhibitory effect of Q-1801 on Raji-Luc tumor growth in B-NDG-hSIRPα mouse model B-luc-GFP Raji cells resuspended in PBS were adjusted to 1×10 5Cells are inoculated into female B-NDG-hSIRPa mice via the tail vein at a concentration of cells / 0.2 mL and a volume of 0.2 mL / mouse. On day 0 after inoculation, a small animal imager is used to observe the tumor inoculation status. On day 4 after inoculation, a small animal imager is used to measure the tumor cell growth status, and tumor-bearing mice with overly strong / weak bioluminescence imaging signals are excluded. Seventy mice with moderate tumor imaging signals are selected and randomly assigned into seven groups of 10 mice each. The average imaging signal of each group is approximately 2.92E+06 p / sec. The day of grouping is designated as D0. On the day of grouping, administration is initiated according to the experimental design, and the administration volume is 10 μL / g. Imaging and dosing are scheduled on the same day, and the interval between imaging and dosing exceeds 4 hours. The detailed administration method, dosage, and administration route are shown in Table 24 below.

[0203]

Table 25

[0204] After the start of drug administration, the status of the mice is carefully observed daily, and the mice are imaged twice a week using a small animal bioluminescence imager to obtain imaging signal diagrams and signal intensities. After the last drug administration, the body weight of the experimental animals and the tumor growth status (detected and recorded by a small animal imager) are continuously observed for 3 days, and then the mice are euthanized.

[0205] Data analysis is performed using the following analysis method: TGI(%) = (1 - TR / CR) × 100%, where TR and CR are the relative tumor imaging signal sizes (R) of the treatment group and the control group at a specific time point, respectively, and R = Vt / V0 (V0 is the average value of the imaging signal at the time of grouping, and Vt is the average value of the imaging signal at each measurement time after treatment).

[0206] The tumor inhibition rate (TGI) was calculated based on the imaging signal intensity, and the results are shown in Table 25 and Figure 54, which is a graph showing the trend of tumor growth after dosing. The body weights of the mice in each group and the tumor imaging signal intensity of each individual mouse in each group after drug administration are shown in Figures 55 to 58. Figure 55 is a graph showing the trend of changes in the body weight of the animals after drug administration. Figure 56 is a mouse bioluminescence imaging photograph on day 0 after grouping. Figure 57 is a mouse bioluminescence imaging photograph on day 7 after grouping. Figure 58 is a mouse bioluminescence imaging photograph on day 14 after grouping.

[0207]

Table 26

[0208] Note: a: Mean ± standard error, b: Statistical comparison of the tumor imaging signal intensity between the dosing group and the solvent control group on day 14 after grouping and dosing, t-test. (**P < 0.01, ****P < 0.0001).

[0209] Experimental results: After 14 days of grouping and dosing, compared with the PBS control group, the experimental groups including 1H9, CHO71, CHO44, and rituximab all significantly inhibited the growth of the tumor imaging signal intensity. The dosing groups using CHO71 and CHO44 in combination with rituximab respectively showed a more significant inhibitory effect on the growth of the tumor imaging signal intensity than the single-dose groups (P < 0.0001 and P < 0.0001). The body weight of the mice during the dosing process did not decrease significantly, indicating that the antibody molecule did not cause obvious toxic side effects to the mice.

[0210] Example 26: Inhibition of the in vivo growth of the MC38-hPD-L1 tumor model in C57BL / 6-hPD-L1 mice Objective: Through the in vivo efficacy test of the MC38-hPD-L1 tumor model in C57BL / 6-hPD-L1 mice, evaluate the inhibitory activity of anti-human SIRPα antibody against tumor growth.

[0211] Digest the mouse colon cancer cells MC38-hPD-L1 in the logarithmic growth phase, remove the culture medium, wash the cells twice with PBS, then count the cells and inoculate them subcutaneously on the right side of C57BL / 6-hPD-L1 mice. Each mouse was inoculated with 5×10 5 / 100 μL of tumor cells. When the average tumor volume grew to about 50 mm 3 , the mice were randomly grouped into 10 mice per group. The day of grouping was defined as D0. On the day of grouping, administration was started according to the experimental protocol design, and the administration volume was 10 μL / g. The detailed administration method, dosage and administration route are shown in Table 26 below.

[0212]

Table 27

[0213] After the start of drug administration, the body weight and tumor volume of the mice were measured twice a week. Tumor volume calculation formula: Tumor volume (mm 3 ) = 1 / 2×(a×b 2 ) (where a represents the long diameter of the tumor and b represents the short diameter of the tumor). The experiment was stopped 1 week after the last drug administration, the mice were euthanized, the tumors were weighed and photographed.

[0214] Select the following analysis method to perform data analysis. The relative tumor growth rate, T / C (%), that is, at a specific time point, is the percentage value of the relative tumor volume or tumor weight of the treatment group and the control group. The calculation formula is as follows. T / C% = TRTV / CRTV×100% (TRTV: average RTV of the treatment group, CRTV, average RTV of the control group, RTV = Vt / V0, V0 is the tumor volume of the mice at the time of grouping, Vt is the tumor volume of the mice after treatment. Based on the tumor volume, the relative tumor inhibition rate TGI TVCalculate (%) and the calculation formula is as follows. TGITV% = (1 - T / C) × 100% (T and C are the relative tumor volumes (RTV) of the treatment group and the control group at a specific time point respectively), and the tumor growth inhibition rate TGI based on the change in tumor weight TW Calculate (%) and the calculation formula is as follows. TGITW% = (1 - TWtreat / TWvehicle) × 100% (TWtreat and TWvehicle are the average values of the tumor weights of the mice in the dosing group and the control group at the end of the experiment respectively).

[0215] Experimental results: The average tumor volume of the mice in the PBS control group was 402.47 mm on the 19th day after drug administration 3 On the 19th day after drug administration, the average tumor volumes of the antibody molecule CHO44-L (7.5 mg / kg), CHO44-M (15 mg / kg), CHO44-H (30 mg / kg), and QP3447 (10 mg / kg) groups were 198.20 mm 3 , 144.21 mm 3 , 92.54 mm 3 and 89.33 mm 3 respectively. Compared with the control group PBS, CHO44-L (7.5 mg / kg, TGI = 58.07%), CHO44-M (15 mg / kg, TGI = 73.19%), CH044-H (30 mg / kg, TGI = 87.94%), and QP3447 (10 mg / kg, TGI = 88.92%) can significantly inhibit tumor growth in a dose-gradient-dependent manner (P < 0.05*, P < 0.05*, P < 0.01**, and P < 0.01**).

[0216] The tumor inhibition rate (TGITV) was calculated based on the mouse tumor volume, and the results are shown in Table 27. The results of the efficacy tests of antibody molecules CHO44-L (7.5 mg / kg), CHO44-M (15 mg / kg), and CHO44-H (30 mg / kg) against the MC38-hPD-L1 tumor model are as shown in Figure 59. Figure 60 shows the grouping of the PBS group, CHO44-L (7.5 mg / kg), CHO44-M (15 mg / kg), and CHO44-H (30 mg / kg) administration groups and the tumor growth curves of each mouse after drug administration. Figure 61 shows the body weight change curves of the mice in each group after drug administration in the MC38-hPD-L1 colon cancer tumor model.

[0217]

Table 28

[0218] Note: a: Data are presented as the mean, b: Compared with the G1 group, an independent samples t-test was used, *: P < 0.05, **: P < 0.01.

[0219] Through the in vivo efficacy test of the MC38-hPD-L1 tumor model in C57BL / 6-hPD-L1 mice, it was found that both antibody molecules CHO44 and QP3447 can significantly inhibit tumor growth, and the inhibitory effect of CHO44 on tumor growth is dose-dependent, and the mouse body weight does not decrease significantly during administration, indicating that the antibody molecules have no obvious toxic side effects on mice.

[0220] Example 27: Inhibition of in vivo growth of CT26-hPD-L1&hCD47 tumors in BALB / c-hPD-1&hSIRPα mice Experimental purpose: To evaluate the inhibitory activity of anti-human SIRPα antibody against CT26 tumor growth through the in vivo efficacy test of CT26-hPD-L1&hCD47 tumors in BALB / c-hPD-1&hSIRPα mice.

[0221] Experimental steps: Logarithmic growth phase mouse colon cancer cells CT26-hPD-L1&hCD47 were digested, the culture medium was removed, and the cells were washed twice with PBS, after which the cells were counted and inoculated subcutaneously into the right flank of BALB / c-hPD-1&hSIRPα transgenic mice, each mouse was inoculated with 1.5×106 / 100μL of tumor cells, and the average tumor volume was about 40mm 3 When the mice reached the age of 10 days, they were randomly divided into groups of 6 mice each, and the day of grouping was defined as D0. On the day of grouping, administration was started according to the experimental protocol design, and the administration volume was 10 μL / g. The detailed administration method, dosage and administration route are shown in Table 28.

[0222] [Table 29]

[0223] After starting drug administration, the body weight and tumor volume of the mice were measured twice a week. Tumor volume was calculated using the following formula: tumor volume (mm 3 )=1 / 2×(a×b 2 ) (where a represents the long diameter of the tumor, and b represents the short diameter of the tumor). After completion of the test, the experiment is stopped, the mice are euthanized, and the tumors are removed, weighed, and photographed.

[0224] The following analysis method was selected to perform data analysis: relative tumor growth rate, T / C (%), i.e., the percentage value of the relative tumor volume or tumor weight of the treatment group and the control group at a specific time point. The calculation formula is as follows: T / C%=TRTV / CRTV×100% (TRTV: average RTV of the treatment group, CRTV, average RTV of the control group, RTV=Vt / V0, V0 is the tumor volume of the mice at the time of grouping, Vt is the tumor volume of the mice after treatment, and the relative tumor inhibition rate TGI was calculated based on the tumor volume. TV The TGI (%) was calculated using the following formula: TV %=(1-T / C)×100% (T and C are the relative tumor volumes (RTVs) of the treatment and control groups, respectively, at a particular time point), tumor inhibition rate TGI based on the change in tumor weight TWCalculate (%) and the calculation formula is as follows. TGI TW % = (1 - TW treat / TW vehicle ) × 100% (TW treat and TW vehicle are the average values of the tumor weights of the mice in the administration group and the control group at the end of the experiment, respectively).

[0225] Experimental results: The average tumor volume of the mice in the PBS control group was 841.5 mm on the 30th day after drug administration 3 . On the 30th day after drug administration, the average tumor volumes of the antibody molecule CHO44-L (15 mg / kg) and CHO44-H (30 mg / kg) groups were 158.3 mm 3 , 128.1 mm 3 respectively. Calculate the tumor growth inhibition rate (TGI TV ) based on the tumor volume of the mice. Compared with the control group PBS, CHO44-L (15 mg / kg, TGI = 80.85%), CH044-H (30 mg / kg, TGI = 83.9%) all significantly inhibited tumor growth (P < 0.05*, P < 0.05*, P < 0.01**, and P < 0.01**). Refer to Table 29 and Figure 62. Figure 63 is the grouping of the PBS group, CHO44-L (15 mg / kg), and CHO44-H (30 mg / kg) administration groups and the tumor growth curves of each mouse after drug administration. Here, in the CHO-L group, tumors disappeared in 3 mice on D14 days (2 mice) and D23 days (1 mouse) respectively, and in the CHO44-H group, tumors disappeared in 4 mice on D12 days (2 mice), D14 days (1 mouse), and D21 days (1 mouse) respectively. Figure 64 is the body weight change curve of each group of mice after drug administration in the CT26-hPD-L1&hCD47 colon cancer tumor model.

[0226]

Table 30

[0227] In the pharmacodynamic evaluation of BALB / c-hPD1 / hSIRPα mice subcutaneously inoculated with CT26-hPDL1hCD47, mice with tumor regression appeared. For mice with tumor regression, CT26-hPDL1&hCD47 was re-inoculated to evaluate the tumor growth status.

[0228] For the cell treatment in the test, the weighing of tumor volume, and the calculation method of TGI, all refer to the in vivo pharmacodynamic evaluation method at the time of the first vaccination. The number of cells inoculated is the same as the number of the first inoculation, the inoculation position is on the left side opposite to the first inoculation position, and the tumor volume is measured twice a week.

[0229] The results show that after the re-inoculation of cells, the average tumor volume grows to 113.56 mm on the 14th day after inoculation in the PBS group. 3 The tumors in the antibody molecule CHO44-L (15 mg / kg), CHO44-H (30 mg / kg) groups do not grow on the 14th day after re-inoculation. For the results, refer to Figure 65, and for the growth curves of mice in each group and the body weights of mice, refer to Figures 66 and 67. The results indicate that mice develop immune memory during the first treatment, and when tumors regress and are re-inoculated, the tumors do not grow.

[0230] Example 28: Pharmacodynamic evaluation in a PBMC reconstitution model of female NCG mice subcutaneously transplanted with the non-small cell lung cancer HCC827 cell line Human non-small cell lung cancer HCC827 cells are subcutaneously inoculated into female NCG mice at 3.0E+06 cells / 100 μl. The average tumor volume is 100 mm 3When it reaches, it is defined as D0 day, and 5.5E+06 cells / mouse of PBMC are inoculated intraperitoneally on D0 day. Seven days later (D7), according to the tumor volume, eight mice in each group are randomly divided into four groups: G1 / PBS, G2 / CHO44-10 mg / kg, G3 / CHO44-25 mg / kg, G4 / CHO71-8 mg / kg + QP3447-4 mg / kg. After the start of drug administration, the body weight of the mice is measured twice a week and the size of the tumor is measured 2-3 times a week. Based on the statistical analysis of the tumor volume data on D24 day, compared with the control group PBS, the G2 / CHO44-10 mg / kg, G3 / CHO44-25 mg / kg, and G4 / CHO71-8 mg / kg + QP3447-4 mg / kg groups all have significant inhibitory activity against tumor growth, and CHO44 shows dose gradient dependence. The TGI is G2: 72.98% (p < 0.0001), G3: 86.28% (p < 0.0001), G4: 58.02% (p = 0.0023) respectively (Figure 68, Table 30). Refer to Figure 69 for the body weight changes during the administration period.

[0231]

Table 31

[0232] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be understood that the present invention is not limited to the above description. Those skilled in the art will be able to see various modifications and changes to the present invention after reading the above content. Therefore, the protection scope of the present invention should be limited by the appended patent claims.

Claims

**Claim 1** A bispecific antibody or an antigen-binding fragment thereof that targets SIRPα and PD-L1, comprising an SIRPα-binding domain and a PD-L1-binding domain, wherein the SIRPα-binding domain comprises a heavy-chain variable region and a light-chain variable region, the heavy-chain variable region comprises VHCDR1, VHCDR2, and VHCDR3 whose amino acid sequences are shown in SEQ ID NOs: 3, 4, and 5 respectively, and the light-chain variable region comprises VLCDR1, VLCDR2, and VLCDR3 whose amino acid sequences are shown in SEQ ID NOs: 37, 38, and 9 respectively, the PD-L1-binding domain comprises a VHH fragment, and the VHH fragment comprises CDR1, CDR2, and CDR3 whose amino acid sequences are shown in SEQ ID NOs: 63, 64, and 65 respectively. The bispecific antibody or an antigen-binding fragment thereof that targets SIRPα and PD-L1 is characterized by this. **Claim 2** The sequence of the heavy-chain variable region of the SIRPα-binding domain is as shown in SEQ ID NO: 17, or has at least 85% sequence identity therewith, or the sequence of the light-chain variable region of the SIRPα-binding domain is selected from SEQ ID NO: 18, or has at least 85% sequence identity therewith, characterized by the bispecific antibody or an antigen-binding fragment thereof that targets SIRPα and PD-L1 according to Claim 1. **Claim 3** The sequence of the VHH fragment is as shown in SEQ ID NO: 62, or has at least 85% sequence identity therewith, characterized by the bispecific antibody or an antigen-binding fragment thereof that targets SIRPα and PD-L1 according to Claim 1. **Claim 4** The bispecific antibody or an antigen-binding fragment thereof further comprises a heavy-chain constant region selected from human IgG1, IgG2, IgG3, or IgG4 or a variant thereof, and a light-chain constant region selected from human κ chain, λ chain, or a variant thereof, characterized by the bispecific antibody or an antigen-binding fragment thereof that targets SIRPα and PD-L1 according to Claim 1. **Claim 5** The heavy-chain constant region comprises an Fc fragment or a variant thereof, and the variant of the Fc fragment is derived from IgG1 and, according to EU count, comprises mutation sites: L234A, L235A, K338A, characterized by The bispecific antibody or antigen-binding fragment thereof targeting SIRPα and PD-L1 according to claim 4.

6. The bispecific antibody or antigen-binding fragment thereof comprises a first polypeptide chain and a second polypeptide chain, The first polypeptide chain comprises the heavy chain variable region of the SIRPα binding domain, the heavy chain constant region, and the VHH fragment, and the VHH fragment is fused to the N-terminus of the heavy chain variable region of the SIRPα binding domain, or the VHH fragment is fused to the C-terminus of the heavy chain constant region, The second polypeptide chain is characterized by comprising the light chain variable region of the SIRPα binding domain and the light chain constant region The bispecific antibody or antigen-binding fragment thereof targeting SIRPα and PD-L1 according to claim 4.

7. The bispecific antibody or antigen-binding fragment thereof comprises a first polypeptide chain and a second polypeptide chain, The first polypeptide chain comprises the heavy chain variable region of the SIRPα binding domain and the heavy chain constant region, The second polypeptide chain is characterized by comprising the light chain variable region of the SIRPα binding domain, the light chain constant region, and the VHH fragment, and the VHH fragment is fused to the N-terminus of the light chain variable region of the SIRPα binding domain The bispecific antibody or antigen-binding fragment thereof targeting SIRPα and PD-L1 according to claim 4.

8. The bispecific antibody or antigen-binding fragment thereof is characterized by having a symmetric structure comprising two of the first polypeptide chains and two of the second polypeptide chains The bispecific antibody or antigen-binding fragment thereof targeting SIRPα and PD-L1 according to claim 6 or 7.

9. The bispecific antibody or antigen-binding fragment thereof further comprises a linker sequence, and preferably, the linker sequence is selected from (GGGGGS)n, where n is an integer from 1 to 4 The bispecific antibody or antigen-binding fragment thereof targeting SIRPα and PD-L1 according to claim 6 or 7.

10. The amino acid sequence of the first polypeptide chain is as shown in any of SEQ ID NO: 66, 26, 69, 84, 85, or the amino acid sequence of the second polypeptide chain is as shown in any of SEQ ID NO: 67, 68, 82, 83 A bispecific antibody targeting SIRPα and PD-L1 or an antigen-binding fragment thereof according to claim 6 or 7.

11. The amino acid sequence of the first polypeptide chain is as shown in SEQ ID NO: 66, and the amino acid sequence of the second polypeptide chain is as shown in SEQ ID NO: 67, characterized in that A bispecific antibody targeting SIRPα and PD-L1 or an antigen-binding fragment thereof according to claim 10.

12. A drug, characterized by comprising a bispecific antibody targeting SIRPα and PD-L1 or an antigen-binding fragment thereof according to any one of claims 1 to 11.

13. characterized in that the drug further comprises one or more other cancer therapeutic agents The drug according to claim 12.

14. A nucleic acid molecule, characterized by encoding a bispecific antibody targeting SIRPα and PD-L1 or an antigen-binding fragment thereof according to any one of claims 1 to 11.

15. A vector, characterized by comprising the nucleic acid molecule according to claim 14.

16. A host cell transformed with the vector according to claim 15.

17. Use of a bispecific antibody targeting SIRPα and PD-L1 or an antigen-binding fragment thereof according to any one of claims 1 to 11 in the preparation of a drug for inhibiting or treating a disease, disorder or condition.

18. The disease, disorder or condition includes cancer, solid tumors, chronic infections, inflammatory diseases, multiple sclerosis, autoimmune diseases, neurological disorders, brain injuries, nerve injuries, polycythemia, hemochromatosis, trauma, septic shock, fibrosis, atherosclerosis, obesity, type II diabetes, graft dysfunction or arthritis. The use according to claim 17.

19. The cancer is selected from anal cancer, appendiceal cancer, astrocytoma, basal cell cancer, gallbladder cancer, gastric cancer, lung cancer, bronchial cancer, bone cancer, hepatobiliary duct cancer, pancreatic cancer, breast cancer, liver cancer, ovarian cancer, testicular cancer, kidney cancer, renal pelvis ureteral cancer, salivary gland cancer, small intestine cancer, urethral cancer, bladder cancer, head and neck cancer, spinal cancer, brain cancer, cervical cancer, uterine cancer, endometrial cancer, colon cancer, colorectal cancer, rectal cancer, esophageal cancer, gastrointestinal cancer, skin cancer, prostate cancer, pituitary cancer, vaginal cancer, thyroid cancer, laryngeal cancer, glioblastoma, melanoma, myelodysplastic syndrome, sarcoma, teratoma, chronic lymphocytic leukemia (CLL), chronic myelogenous leukemia (CML), acute lymphocytic leukemia (ALL), acute myelogenous leukemia (AML), Hodgkin lymphoma, non-Hodgkin lymphoma, multiple myeloma, T-cell or B-cell lymphoma, gastrointestinal stromal tumor, soft tissue tumor, hepatocellular cancer or adenocarcinoma The use according to claim 18.

20. The drug is characterized by being used in combination with one or more other drugs The use according to claim 17.

21. The other drug is characterized by containing rituximab The use according to claim 20.

Images