A bispecific antibody targeting CD47 and HER2 and its preparation method and application

Abstract

The present invention belongs to the field of antibodies, specifically to a bispecific antibody targeting CD47 and HER2, its preparation method, and its application. The present invention utilizes genetic engineering techniques to mutate the Fc regions of anti-human CD47 and anti-human HER2 monoclonal antibodies, enabling in vitro assembly of the "1+1" bispecific antibody D0604T through Fab arm exchange. The bispecific antibody D0604T of the present invention effectively recognizes and binds to CD47 and HER2 on tumor cells while not causing significant coagulation or binding to human red blood cells at high concentrations.

Description

Technical Field

[0001] The present invention belongs to the field of antibodies, and specifically relates to a bispecific antibody targeting CD47 and HER2, and a preparation method and application thereof. Background Art

[0002] Cancer is one of the major diseases that threaten human life and health, and is also one of the most worrying issues in the medical field. Currently, the commonly used methods for treating cancer include surgery, chemotherapy, radiotherapy, and immunotherapy. Surgery, chemotherapy, and radiotherapy were discovered early and have been used to this day, improving patients' survival. However, as treatment time increases, patients are prone to drug resistance and recurrence, posing a huge challenge to the clinical treatment of cancer. Therefore, finding new treatment strategies and targets has become one of the current hot topics in cancer research. The emergence of tumor immunotherapy has reversed this situation. Among them, antibody drugs are a major invention in this field. They have the advantages of strong targeting, high specificity, and few toxic side effects, and are currently the mainstream drugs for treating cancer.

[0003] Immune checkpoint molecules play a key role in regulating anti-tumor immune responses. Tumor cells often exploit these pathways to suppress and evade the immune system. Targeting PD-1, CTLA4, and LAG3 has been approved by the FDA for the treatment of various solid tumors. While their effectiveness has been demonstrated, sustained or complete remissions are achieved in only a small percentage of patients. Macrophages are the first line of defense in immune surveillance, maintaining a balance between "eat me" and "don't eat me" signals. Multiple immune checkpoints, including SIRPα, Siglec-10, LILRB1, and PD-1, are present on the surface of macrophages. These checkpoints rely on phosphorylation of immunoreceptor tyrosine inhibitory motifs (ITIMs) to recruit downstream SH2 domain-containing protein tyrosine phosphatases 1 or 2 (SHP-1, SHP-2), which transmit inhibitory signals and influence cytoskeletal rearrangements, thereby inhibiting macrophage phagocytosis. Among these immune checkpoints, besides PD-1, therapeutic strategies targeting cluster of differentiation 47 (CD47) are the most mature.

[0004] CD47 is a ubiquitously expressed transmembrane glycoprotein, also known as integrin-associated protein (IAP). It is a glycosylated transmembrane protein widely expressed on the surface of various cells. CD47 is upregulated on the surface of various tumors and interacts with its ligand, signal regulatory protein α (SIRPα), on the surface of macrophages, mimicking the self-protection mechanism of normal cells and enabling immune escape.

[0005] Using antibodies to block the CD47-SIRPα axis can restore macrophage recognition and phagocytosis of tumor cells. According to public data, including preclinical projects, there are currently more than 100 CD47 projects under development worldwide, of which antibody drugs account for the majority. However, two major challenges limit the development of CD47 target drugs: (1) On-target toxicity: In addition to tumor cells, CD47 is widely expressed on normal cells (especially red blood cells), which can easily cause severe blood toxicity at therapeutic doses; (2) Limited efficacy in solid tumors: The response rate of CD47 monoclonal antibodies in solid tumors is significantly lower than that in hematological tumors. Therefore, the core of CD47 antibody drug development should be to find a balance between effectiveness and toxicity, while protecting red blood cells and killing tumor cells to the greatest extent.

[0006] Human epidermal growth factor receptor 2 (HER2) is a receptor-like transmembrane protein that is activated in some human tumors. The gene encoding the HER2 protein is amplified in 20%-25% of breast cancer patients. HER2-targeted antibodies, including Herceptin and Perjeta, are currently the mainstay of targeted therapy for HER2-overexpressing cancers. These antibodies, by targeting HER2, induce antibody-dependent cell-mediated cytotoxicity (ADCC) and classical complement activation, thereby inhibiting tumor cell growth, division, and spread. Studies have shown that there is a potential correlation between the expression and function of CD47 and HER2. When CD47 monoclonal antibodies were used in combination with HER2 monoclonal antibodies, the results showed that the ADCP effect mediated by CD47 monoclonal antibodies could enhance the anti-tumor effect of trastuzumab (Tsao LC, et al. CD47 blockadeaugmentation of trastuzumab antitumor efficacy dependent on antibody-dependent cellular phagocytosis. JCI Insight. 2019 Dec 19;4(24):e131882.).

[0007] Based on this, the present invention aims to develop a bispecific antibody targeting CD47 and HER2, which enhances the specific tumor targeting of the CD47 antibody and improves the red blood cell safety by introducing the HER2 target; and the synergistic effect of the CD47 and HER2 dual targets enhances the anti-tumor activity, thereby solving the problem of the inability to achieve a balance between the safety and efficacy of the CD47 monoclonal antibody. Summary of the Invention

[0008] Purpose of the Invention: The present invention provides a bispecific antibody, D0604T, that targets CD47 and HER2. This bispecific antibody targets HER2 with relatively high affinity and CD47 with relatively low affinity, thereby improving tumor-specific targeting and increasing red blood cell safety. Furthermore, the synergistic effect of the dual CD47 and HER2 targets enhances anti-tumor activity, addressing the current issue of CD47 monoclonal antibodies that struggle to achieve a balance between safety and efficacy.

[0009] Technical solution: A bispecific antibody targeting CD47 and HER2, the bispecific antibody being D0604T. D0604T comprises four polypeptide chains. The first chain has the amino acid sequence shown in SEQ ID NO.7 and is composed of the heavy chain variable region of the CD47 antibody shown in SEQ ID NO.3 and the mutated heavy chain constant region shown in SEQ ID NO.1; the second chain has the amino acid sequence shown in SEQ ID NO.8 and is composed of the heavy chain variable region of the HER2 antibody shown in SEQ ID NO.5 and the mutated heavy chain constant region shown in SEQ ID NO.2; the third chain has the amino acid sequence shown in SEQ ID NO.9 and is composed of the light chain variable region of the CD47 antibody shown in SEQ ID NO.4 and the light chain constant region shown in SEQ ID NO.11; and the fourth chain has the amino acid sequence shown in SEQ ID NO.10 and is composed of the light chain variable region of the HER2 antibody shown in SEQ ID NO.6 and the light chain constant region shown in SEQ ID NO.11.

[0010] A method for preparing a bispecific antibody targeting CD47 and HER2 comprises utilizing Duobody technology derived from patent WO2011131746 to perform Fc region mutation engineering on an IgG1 CD47 monoclonal antibody and an IgG1 HER2 monoclonal antibody. The heavy chain constant region sequence of the CD47 monoclonal antibody comprises the amino acid sequence shown in SEQ ID NO.1, with the amino acid phenylalanine at position 405 of the heavy chain mutated to leucine according to EU numbering; the heavy chain constant region sequence of the HER2 monoclonal antibody comprises the amino acid sequence shown in SEQ ID NO.2, with the amino acid lysine at position 409 of the heavy chain mutated to arginine according to EU numbering. The CD47 monoclonal antibody and the HER2 monoclonal antibody with Fc region mutation engineering are mixed in an equimolar ratio and can be self-assembled under in vitro reducing conditions to form a "1+1" bispecific antibody D0604T that simultaneously targets CD47 and HER2.

[0011] The CD47 monoclonal antibody molecule D0604 sequence is derived from ZL202211709826.6, and is composed of a heavy chain and a light chain. The heavy chain variable region sequence includes the amino acid sequence shown in SEQ ID NO.3; the light chain variable region sequence includes the amino acid sequence shown in SEQ ID NO.4.

[0012] The HER2 monoclonal antibody molecular sequence is derived from patent US10167342, and is composed of a heavy chain and a light chain. The heavy chain variable region sequence includes the amino acid sequence shown in SEQ ID NO.5; the light chain variable region sequence includes the amino acid sequence shown in SEQ ID NO.6.

[0013] An expression vector comprising the above-mentioned bispecific antibody targeting CD47 and HER2.

[0014] A recombinant cell comprising the above expression vector.

[0015] The present invention also discloses the use of a bispecific antibody targeting CD47 and HER2 in the preparation of a drug for treating cancer.

[0016] Furthermore, the cancer is breast cancer or ovarian cancer.

[0017] EU numbering refers to Edelman GM, Cunningham BA, Gall WE, Gottlieb PD,Rutishauser U, Waxdal MJ. The covalent structure of an entire gammaGimmunoglobulin molecule. Proc Natl Acad Sci US A. 1969 May;63(1):78-85. This numbering method comes from an earlier discovered IgG1 immunoglobulin named EU and is sorted according to the amino acid sequence of the entire EU antibody (starting from the variable region).

[0018] Beneficial effects: Although bispecific antibodies (BsAbs) theoretically have advantages such as simultaneously targeting two antigen epitopes, enhancing specificity, or recruiting immune cells, they are not superior to monoclonal antibodies in all cases. For example:

[0019] 1. Clinical efficacy does not meet expectations:

[0020] A. Improper target combination selection: The design of bispecific antibodies relies on the synergistic effect of two targets. If the biological mechanisms of the targets are unclear or the combination is inappropriate, poor efficacy may result. For example, Amgen's Bintrafusp alfa (targeting PD-L1 and TGF-β) failed in a Phase III lung cancer trial because TGF-β inhibition may promote immunosuppression in the tumor microenvironment, offsetting the efficacy of PD-L1 blockade. Merck's M7824 (with a similar target) showed some efficacy in a Phase II trial for cholangiocarcinoma, but failed to surpass PD-1 monoclonal antibodies in other indications.

[0021] B. Lack of synergistic effect: Although some bispecific antibodies can bind to two targets, they do not produce the expected synergistic effect. For example:

[0022] AbbVie's ABT-165 (targeting VEGF and DLL4): Development was terminated due to limited efficacy and high toxicity in solid tumors.

[0023] 2. Toxicity issues

[0024] A. Off-target toxicity or overactivation of the immune system: Dual antibodies can cause unpredictable toxicities by acting on two pathways simultaneously. For example, the CD19 / CD3 dual antibody (blinatumomab), while approved for leukemia, is associated with severe cytokine release syndrome (CRS) and neurotoxicity, limiting its widespread use. HER2 dual antibodies (such as zanidatamab) can cause cardiotoxicity in some patients (related to the role of HER2 signaling in cardiomyocytes).

[0025] B. Target expression overlap: If two targets are widely co-expressed in normal tissues, this may lead to on-target / off-tumor toxicity. For example, a dual antibody targeting EGFR and c-MET may exacerbate skin and gastrointestinal adverse reactions.

[0026] Examples of current failures include Roche's Vanucizumab (targeting VEGF-A and Ang2): its efficacy in Phase III colorectal cancer was not superior to that of bevacizumab (an anti-VEGF monoclonal antibody), leading to project termination. Sanofi's SAR156597 (targeting IL-4 / IL-13): its Phase II idiopathic pulmonary fibrosis trial failed due to a failure to improve lung function.

[0027] This invention is a bispecific antibody. Using Duobody technology, specific amino acids in the constant regions of the heavy chains of CD47 and HER2 monoclonal antibodies are mutated, and disulfide bonds are then used to construct the anti-CD47 / HER2 bispecific antibody D0604T. As is common knowledge, bispecific antibodies are not inherently superior to monoclonal antibodies; their efficacy and safety must be considered.

[0028] CD47 is widely expressed in normal cells (especially red blood cells), so CD47 antibodies can easily cause severe blood toxicity (hemolytic reaction) at therapeutic doses. How to reduce these toxic side effects is a technical problem that needs to be solved urgently in this field.

[0029] The bispecific antibody D0604T of the present invention effectively binds to hHER2 at the protein level, with comparable binding capacity to bivalent monoclonal antibodies. Bispecific antibody D0604T also effectively binds to hCD47 at the nanomolar level. Even though D0604T's binding capacity is slightly reduced compared to its precursor monoclonal antibody D0604, it still binds to human CD47 in a concentration-dependent manner. Clinically tested control CD47 monoclonal antibodies IBI-188 and Hu5F9 both cause significant erythrocyte agglutination at concentrations above 12 nM, with significant erythrocyte lysis at high concentrations. TJC4 causes mild erythrocyte agglutination at concentrations above 12 nM. However, D0604, D0604T, and the most clinically advanced CD47 monoclonal antibody AK117 do not cause erythrocyte agglutination at 3000 nM, demonstrating the high erythrocyte safety of bispecific antibodies. Further examination of the antibodies' erythrocyte binding ability revealed that Hu5F9 and IBI-188 readily bound to erythrocytes, with binding levels significantly higher than those of the other antibodies. AK117 and D0604, on the other hand, bound to erythrocytes at concentrations above 25 nM, with binding activity significantly reduced at lower concentrations. Compared to the monoclonal antibody D0604, the bispecific antibody D0604T showed virtually no erythrocyte binding at 200 nM, further reducing the monoclonal antibody's erythrocyte binding capacity and demonstrating superior erythrocyte safety. Compared to the monoclonal antibody D0604, the bispecific antibody D0604T exhibited significantly enhanced SIRPα blocking activity at low concentrations, effectively blocking the "don't eat me" signal on tumor cells.

[0030] In summary, the bispecific antibody D0604T of the present invention not only achieves dual targeting function compared with CD47 and HER2 monoclonal antibodies, but also has better anti-tumor effect than monoclonal antibody D0604 (CD47 monoclonal antibody), among which EC 50 (monoclonal antibody D0604) is 10 times that of D0604T, so the bispecific antibody D0604T is safer and more effective. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Schematic diagram of the molecular structure of the bispecific antibody molecule D0604T and its precursor monoclonal antibody, where A: CD47 precursor monoclonal antibody D0604, HER2 precursor monoclonal antibody; B: bispecific antibody D0604T.

[0032] Figure 2 The molecular weight of the bispecific antibody protein was determined by SDS-PAGE.

[0033] Figure 3HIC-HPLC analysis of the bispecific antibody assembly efficiency, wherein A: HIC-HPLC analysis of D0604; B: HIC-HPLC analysis of HER2 antibody; C: HIC-HPLC analysis of D0604T.

[0034] Figure 4 ELISA was used to detect the binding ability of the bispecific antibody to the target antigen human CD47 at the protein level.

[0035] Figure 5 ELISA was used to detect the binding ability of the bispecific antibody to the target antigen human HER2 at the protein level.

[0036] Figure 6 Flow cytometry was used to detect the binding ability of bispecific antibodies to double-positive cells endogenously expressing CD47 and HER2. A: Flow cytometry comparison of the binding of bispecific antibodies and monoclonal antibodies to SKOV3 cells; B: Flow cytometry comparison of the binding of bispecific antibodies and monoclonal antibodies to SKBR3 cells.

[0037] Figure 7 The results are shown for the bispecific antibody inducing hemagglutination at different concentrations.

[0038] Figure 8 The binding ability of the bispecific antibody to red blood cells was determined by flow cytometry.

[0039] Figure 9 Flow cytometry was used to detect the SIRPα blocking activity of the bispecific antibody against ovarian cancer cells SKOV3 that endogenously express CD47 and HER2. DETAILED DESCRIPTION

[0040] Meanings of abbreviations in the following examples or figures: D0604 is a CD47 antibody; D0604T is a bispecific antibody; anti-HER2 is a HER2 antibody.

[0041] AK117 is a humanized IgG4 monoclonal antibody developed by Akeso, Inc. It primarily targets CD47 and can bind to CD47 expressed on tumor cells, blocking the interaction between CD47 and its receptor SIRPα without causing the side effect of red blood cell coagulation. However, it still retains a certain amount of red blood cell binding, posing a risk to red blood cell safety.

[0042] Hu5F9, IBI-188, and TJC4 are also different CD47 monoclonal antibodies that can bind to CD47 expressed on tumor cells and block the interaction between CD47 and its receptor SIRPα.

[0043] Example 1, Preparation of bispecific antibodies:

[0044] Construction of expression vectors for light and heavy chains of bispecific antibody precursor monoclonal antibody molecules and their eukaryotic expression:

[0045] Upstream and downstream primers were designed, and PCR was used to obtain the variable regions of CD47 and HER2 antibodies, the full-length light chains of CD47 and HER2, and the constant region of IgG1 heavy chain, respectively. Primers for mutations at positions 405 and 409 of the constant region (according to EU numbering) were designed, and PCR was used to obtain the constant regions IgG1-F405L (Phe at position 405 mutated to Leu) and IgG1-K409R (Lys at position 409 mutated to Arg) containing Duobody mutations. The heavy chain variable region of CD47 antibody was spliced ​​with the IgG1 heavy chain constant region IgG1-F405L, and the heavy chain variable region of HER2 antibody was spliced ​​with the IgG1 heavy chain constant region IgG1-K409R by overlap-PCR to obtain the full-length heavy chain fragments of the precursor monoclonal antibody (CD47 monoclonal antibody and HER2 monoclonal antibody containing Duobody mutations). EcoRⅠ and NotⅠ The double enzyme digestion expression vector is connected to the full-length fragments of the light and heavy chains of the precursor monoclonal antibody and the enzyme digestion vector respectively through a one-step cloning method to obtain the light and heavy chain expression vectors of the precursor monoclonal antibody.

[0046] The recombinant plasmids encoding the light and heavy chains of the precursor monoclonal antibody were transfected into HEK293 cells at a ratio of 2:1. After culturing for 5 days, the supernatant was collected by centrifugation at 4°C and purified by protein A column to obtain the precursor monoclonal antibody. The structure of the prepared precursor monoclonal antibody is as follows: Figure 1 shown.

[0047] The two precursor monoclonal antibodies were added in an equimolar ratio, and the disulfide bonds of the precursor monoclonal antibodies were reduced with 10 mM TCEP. After the disulfide bonds between the heavy chains were broken, they were re-oxidized with DHAA to allow the precursor half monoclonal antibodies to automatically assemble into a "1+1" heterologous bispecific antibody. The ultrafiltration concentration replacement system was used to obtain a highly pure bispecific antibody named D0604T, whose structure is shown below. Figure 1 As shown in B.

[0048] Table 1: Primer sequences: .

[0049] Example 2, SDS-PAGE identification of bispecific antibodies:

[0050] Prepare 6% SDS-PAGE gel, dilute 2 μg protein sample with PBS buffer to 12 μL, add 3 μL protein loading buffer, mix well and heat in boiling water for 8 min, load the sample, and set the electrophoresis program to constant voltage 80 V and time 100 min.

[0051] Results: As Figure 2As shown, the bispecific antibody bands obtained in Example 1 are correctly positioned and have high purity.

[0052] Example 3, HIC-HPLC identification of bispecific antibody assembly efficiency:

[0053] The hydrophobicity of the bispecific antibody was tested using HIC-HPLC (hydrophobic intercalation chromatography). The assembly efficiency of the bispecific antibody was determined by the difference in hydrophobicity of the precursor monoclonal antibody. The specific experimental conditions were as follows: chromatograph model Fulli LC5190, antibody concentration of 2 mg / mL, injection volume of 20 μL, flow rate of 0.8 mL / min, gradient elution (elution conditions are shown in Table 2), detector wavelength of 280 nm, column temperature of 30°C, and chromatographic column model TSKgel Butyl-NPR.

[0054] Results: See Figure 3 As shown in Table 3, the retention times of the precursor monoclonal antibody D0604 and HER2 antibody were 14.666 min and 9.435 min, respectively. The in vitro assembled "1+1" type bispecific antibody theoretically has a hydrophobicity between the two precursor monoclonal antibodies, with a retention time of 12.478 min. The main peak of the bispecific antibody accounts for 94%, and the bispecific antibody has a high purity.

[0055] Table 2 HIC-HPLC elution conditions: ;

[0056] Table 3 HIC-HPLC test results of bispecific antibodies: .

[0057] Example 4, ELISA detection of the binding ability of the bispecific antibody to the target antigen human CD47 at the protein level:

[0058] Binding of the bispecific antibody to recombinant human CD47 antigen was assessed using ELISA. The mFc-tagged hCD47 recombinant protein was diluted to 1 μg / ml using antigen coating buffer and added to ELISA strips at a volume of 50 μl / well. The wells were coated in duplicate and incubated overnight at 4°C. The strips were washed three times with PBST and blocked with 5% skim milk at 37°C for 2 h. The strips were washed three times with PBST, and serially diluted bispecific antibody, isotype control antibody, and control antibodies D0604 and Hu5F9 (maximum concentration 200 nM, 4-fold serial dilutions, 12 concentrations total) were added and incubated at 37°C for 2 h. The strips were washed three times with PBST, and HRP-conjugated goat anti-human IgG secondary antibody was added. The strips were incubated at 37°C for 1 h, washed six times with PBST, and TMB colorimetric solution was added. The strips were incubated in the dark for 15 min, and the colorimetric reaction was terminated. The absorbance at 450 nm and 630 nm was read on a microplate reader, and the EC50 was calculated.

[0059] Results: As Figure 4 As shown in Table 4, the bispecific antibody can effectively bind to hCD47 protein at the protein level. Since it is at the nM level, even though the binding ability of D0604T is slightly reduced compared with the precursor monoclonal antibody D0604, it can still bind to human CD47 in a concentration-dependent manner.

[0060] Table 4 EC50 values ​​of bispecific antibodies binding to hCD47 protein: .

[0061] Example 5, ELISA detection of the binding ability of bispecific antibodies to the target antigen human HER2 at the protein level:

[0062] Binding of the bispecific antibody to recombinant human HER2 antigen was assessed using ELISA. His-tagged hHER2 recombinant protein was diluted to 1 μg / ml using antigen coating buffer and added to enzyme-labeled strips at a volume of 50 μl / well. The wells were set up in duplicate and coated overnight at 4°C. The strips were washed three times with PBST and blocked with 5% skim milk at 37°C for 2 h. The strips were washed three times with PBST, and serially diluted bispecific antibody, isotype control antibody, and HER2 control antibody were added (maximum concentration 100 nM, with a three-fold serial dilution, for a total of 10 concentrations) and incubated at 37°C for 2 h. The strips were washed three times with PBST, and HRP-conjugated goat anti-human IgG secondary antibody was added. The strips were incubated at 37°C for 1 h, washed six times with PBST, and TMB colorimetric solution was added. The strips were incubated in the dark for 15 min, and the colorimetric reaction was terminated. The absorbance at 450 nm and 630 nm was read on a microplate reader, and the EC50 was calculated.

[0063] Results: As Figure 5 As shown in Table 5 , the bispecific antibody can effectively bind to hHER2 protein at the protein level, and its ability to bind to hHER2 is comparable to that of the bivalent monoclonal antibody.

[0064] Table 5 EC50 values ​​of bispecific antibodies binding to hHER2 protein levels: .

[0065] Example 6, flow cytometry detection of the binding ability of bispecific antibodies to CD47 and HER2 double-positive cells:

[0066] 1) Bispecific antibodies and CD47 + HER2 + The binding of ovarian cancer cell line SKOV3 and breast cancer cell line SKBR3 was detected by flow cytometry. SKOV3 cells were cultured in 1640 medium containing 10% FBS, and SKBR3 cells were cultured in McCoy's 5A medium containing 10% FBS at 37°C and 5% CO2.

[0067] 2) Digest the cells in good growth state with trypsin, add fresh culture medium to resuspend and count, and take 2*10 5 The treated cell suspension was centrifuged at 1500 rpm for 3 min at 4°C, and the supernatant was discarded. 100 μl of pre-chilled PBS containing 2% BSA was added and incubated for 20 min.

[0068] 3) Centrifuge at 1500 rpm at 4°C for 3 minutes and discard the supernatant. Add 100 μL of serially diluted bispecific antibody, monoclonal antibody, and isotype control antibody as the primary antibody (maximum concentration is 200 nM, with a 3-fold serial dilution, totaling 10 concentrations). Resuspend the cells and incubate at 4°C for 1 hour to allow antibody binding to the cells. Centrifuge at 1500 rpm at 4°C for 3 minutes and discard the supernatant. Wash once with 500 μL of pre-chilled PBS. The control antibody AK117 was prepared based on the sequence published in patent WO2021043220.

[0069] 4) Add 100 μL of Alexa Fluor 488-labeled goat anti-human IgG antibody (Yisheng Bio, 1:200 dilution) as a secondary antibody, resuspend the cells, and incubate at 4°C for 30 min.

[0070] 5) Centrifuge at 1500 rpm for 3 minutes at 4°C and discard the supernatant. Wash twice with 500 μL of pre-chilled PBS. Resuspend in 250 μL of PBS and analyze using a flow cytometer.

[0071] Results: As Figure 6 As shown, the bispecific antibodies of the present invention can effectively bind to SKOV3 and SKBR3 cells, which endogenously overexpress CD47 and HER2. Compared to the HER2 monoclonal antibody, the EC50 of the bispecific antibody D0604T is slightly lower. Compared to the CD47 monoclonal antibody, the EC50 of the bispecific antibody D0604T is significantly increased, and its maximum MFI is also significantly increased, demonstrating that the bispecific antibody has better tumor cell targeting activity.

[0072] Example 7, Investigation of Hemagglutination Induced by Bispecific Antibodies at Different Concentrations:

[0073] The control antibody Hu5F9 was prepared based on the sequence published in patent US12060423, TJC4 was prepared based on the sequence published in patent WO2022078465, and IBI-188 was prepared based on the sequence published in patent WO2020173431.

[0074] Fresh blood was collected from healthy donors, lymphocytes were separated using lymphocyte separation solution, 500 μl of red blood cells were aspirated after centrifugation, 3 ml of PBS was added to wash the cells, and the cells were centrifuged at 2000 rpm at room temperature for 5 minutes, for a total of three washes; the red blood cells were resuspended with PBS and diluted to a 2% red blood cell suspension, and added to a 96-well plate for red blood cell plating, with 50 μl of cell suspension per well; the highest final incubation concentration of bispecific antibodies and monoclonal antibodies was set to 3000 nM, and they were diluted three times in succession with PBS, for a total of 12 concentration gradients; 50 μl of CD47 antibody of each concentration gradient was added to each well, incubated at 37°C for 4 hours, and the aggregation of red blood cells in each well was observed. The results are as follows Figure 7 As shown, clinical stage control CD47 monoclonal antibodies IBI-188 and Hu5F9 both caused significant erythrocyte agglutination at concentrations above 12 nM, and erythrocyte lysis was obvious at high concentrations. TJC4 caused mild erythrocyte agglutination at concentrations above 12 nM; D0604, D0604T and the CD47 monoclonal antibody AK117 with the fastest clinical progress did not cause erythrocyte agglutination at 3000 nM, indicating that bispecific antibodies have a high erythrocyte safety.

[0075] Example 8, flow cytometry determination of the binding ability of bispecific antibodies to red blood cells:

[0076] 20 μl of 2% red blood cell suspension was aspirated from each tube, and the cells were resuspended in PBS containing 2% BSA, blocked at 4°C for 20 minutes, centrifuged at 1500 rpm at 4°C for 5 minutes, and the supernatant was discarded; the highest final incubation concentration of bispecific antibodies and monoclonal antibodies was set to 200 nM, and diluted continuously 2-fold with PBS, with a total of 8 concentration gradients, and 100 μl of red blood cells were resuspended in each tube and incubated at 4°C for 1 hour; the cells were resuspended and washed once with 500 μl of PBS; the cells were resuspended with 100 μl of diluted Alexa 488-labeled goat anti-human IgG antibody, incubated at 4°C for 30 minutes, centrifuged at 1500 rpm at 4°C for 5 minutes, and the secondary antibody was discarded; the cells were resuspended and washed twice with 500 μl of PBS, and 250 μL of PBS was added to resuspend the cells. After filtering through a 200-mesh cell screen, the cells were detected by flow cytometry.

[0077] Results: As Figure 8 As shown in Table 6, each antibody binds to erythrocytes to varying degrees, exhibiting a dose-dependent pattern. Hu5F9 and IBI-188 highly readily bind to erythrocytes, with binding levels significantly higher than those of the other antibodies. AK117 and D0604 bind to erythrocytes at concentrations above 25 nM, with binding activity significantly decreasing at lower concentrations. Compared to the monoclonal antibody D0604, the bispecific antibody D0604T exhibits virtually no erythrocyte binding at 200 nM, further reducing the monoclonal antibody's erythrocyte binding ability and demonstrating superior erythrocyte safety.

[0078] Table 6 EC50 values ​​of bispecific antibodies binding to human erythrocytes: .

[0079] Example 9, flow cytometry detection of the ability of antibodies to block CD47–SIRPα binding in SKOV3 cells:

[0080] To evaluate the ability of D0604T and its precursor monoclonal antibody D0604 to block the binding of CD47 to SIRPα on the surface of SKOV3 cells, a flow cytometry competition binding assay was performed. SKOV3 cells were cultured in 1640 medium supplemented with 10% FBS at 37°C and 5% CO2. Well-growing cells were trypsinized, resuspended in fresh medium, and counted. 2 × 10 cells were sampled per tube. 5 The treated cell suspension was centrifuged at 1500 rpm for 3 minutes at 4°C, and the supernatant discarded. 100 μL of pre-chilled PBS containing 2% BSA was added and incubated for 20 minutes. SKOV3 cells were incubated with a three-fold serial dilution of the candidate antibody starting at 200 nM, for a total of 10 concentrations, at 4°C for 1 hour. Subsequently, 10 nM recombinant human SIRPα protein was added and incubated at 37°C for 30 minutes. After washing, Alexa Fluor 488-conjugated goat anti-human IgG secondary antibody was added and incubated at 4°C for 30 minutes. Flow cytometry was then used for analysis.

[0081] Results: As Figure 9 As shown in Table 7, compared with the monoclonal antibody D0604, the bispecific antibody D0604T significantly increased the blocking activity against SIRPα by 10 times, and can better block the "don't eat me" signal on tumor cells.

[0082] Table 7 EC50 values ​​of bispecific antibodies blocking SIRPα: .

Claims

1. A bispecific antibody targeting CD47 and HER2, characterized in that: It is composed of four polypeptide chains; wherein: the first chain has the amino acid sequence shown in SEQ ID NO.7, and is composed of the heavy chain variable region of the CD47 antibody shown in SEQ ID NO.3 and the mutated heavy chain constant region shown in SEQ ID NO.1; the second chain has the amino acid sequence shown in SEQ ID NO.8, and is composed of the heavy chain variable region of the HER2 antibody shown in SEQ ID NO.5 and the mutated heavy chain constant region shown in SEQ ID NO.2; the third chain has the amino acid sequence shown in SEQ ID NO.9, and is composed of the light chain variable region of the CD47 antibody shown in SEQ ID NO.4 and the light chain constant region shown in SEQ ID NO.11; the fourth chain has the amino acid sequence shown in SEQ ID NO.10, and is composed of the light chain variable region of the HER2 antibody shown in SEQ ID NO.6 and the light chain constant region shown in SEQ ID NO.

11.

2. The method for preparing a bispecific antibody targeting CD47 and HER2 according to claim 1, characterized in that: Duobody technology is used to perform Fc region mutation modification on CD47 monoclonal antibody and HER2 monoclonal antibody; the heavy chain constant region sequence of the CD47 monoclonal antibody has the amino acid sequence shown in SEQ ID NO.1, wherein the amino acid phenylalanine at position 405 of the heavy chain according to EU numbering is mutated to leucine; the heavy chain constant region sequence of the HER2 monoclonal antibody has the amino acid sequence shown in SEQ ID NO.2, wherein the amino acid lysine at position 409 of the heavy chain according to EU numbering is mutated to arginine; the CD47 monoclonal antibody and HER2 monoclonal antibody with Fc region mutation modification are mixed in an equimolar ratio, and the disulfide bonds between the heavy chains are broken in the presence of a reducing agent to form CD47 half-monoclonal antibody and HER2 half-monoclonal antibody, which are then automatically assembled in the presence of an oxidant to form a heterologous bispecific antibody.

3. An expression vector, characterized in that Encodes the antibody of claim 1.

4. A recombinant cell, characterized in that Comprising the expression vector according to claim 3.

5. Use of a bispecific antibody targeting CD47 and HER2 according to claim 1 in the preparation of a drug for treating cancer; the cancer is breast cancer or ovarian cancer.

Images