ANTI-CD47/ANTI-PD-L1 ANTIBODY AND APPLICATIONS THEREOF

Abstract

Provided are an anti-CD47/anti-PD-L1 antibody, a pharmaceutical composition of the anti-CD47/anti-PD-L1 antibody, and applications thereof. The anti-CD47/anti-PD-L1 antibody provides antitumor activity, is free of significant red blood cell toxicity, and is applicable in preparing an antitumor medicament.

Claims

1. An anti-CD47/anti-PD-L1 antibody comprising an anti-CD47 antibody or antigen-binding fragment thereof and an anti-PD-L1 antibody or antigen-binding fragment thereof, wherein: the anti-CD47 antibody or antigen-binding fragment thereof comprises a first heavy chain variable region and a first light chain variable region, wherein: (1) the first heavy chain variable region comprises H1CDR1, H1CDR2 and H1CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 4, 5 and 6, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 4, 5 and 6, respectively; and (2) the first light chain variable region comprises L1CDR1, L1CDR2 and L1CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 15, 16 and 17, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 15, 16 and 17, respectively.

2. The anti-CD47/anti-PD-L1 antibody according to claim 1, wherein the anti-PD-L1 antibody or antigen-binding fragment thereof comprises a second heavy chain variable region and a second light chain variable region, wherein: (1) the second heavy chain variable region comprises H2CDR1, H2CDR2 and H2CDR3 selected from the group consisting of: (A1) amino acid sequences set forth in SEQ ID NOs: 75, 76 and 77; (A2) amino acid sequences set forth in SEQ ID NOs: 81, 82 and 83; (A3) amino acid sequences set forth in SEQ ID NOs: 87, 88 and 89; (A4) amino acid sequences set forth in SEQ ID NOs: 93, 94 and 95; and (A5) amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in (A1), (A2), (A3) or (A4); and (2) the second light chain variable region comprises L2CDR1, L2CDR2 and L2CDR3 selected from the group consisting of: (A6) amino acid sequences set forth in SEQ ID NOs: 78, 79 and 80; (A7) amino acid sequences set forth in SEQ ID NOs: 84, 85 and 86; (A8) amino acid sequences set forth in SEQ ID NOs: 90, 91 and 92; (A9) amino acid sequences set forth in SEQ ID NOs: 96, 97 and 98; and (A10) amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in (A6), (A7), (A8) or (A9).

3. The anti-CD47/anti-PD-L1 antibody according to claim 2, wherein in the anti-PD-L1 antibody or antigen-binding fragment thereof: the second heavy chain variable region comprises H2CDR1, H2CDR2 and H2CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 75, 76 and 77, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 75, 76 and 77, respectively, and the second light chain variable region comprises L2CDR1, L2CDR2 and L2CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 78, 79 and 80, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 78, 79 and 80, respectively; the second heavy chain variable region comprises H2CDR1, H2CDR2 and H2CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 87, 88 and 89, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 87, 88 and 89, respectively, and the second light chain variable region comprises L2CDR1, L2CDR2 and L2CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 90, 91 and 92, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 90, 91 and 92, respectively; the second heavy chain variable region comprises H2CDR1, H2CDR2 and H2CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 81, 82 and 83, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 81, 82 and 83, respectively, and the second light chain variable region comprises L2CDR1, L2CDR2 and L2CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 84, 85 and 86, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 84, 85 and 86, respectively; or the second heavy chain variable region comprises H2CDR1, H2CDR2 and H2CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 93, 94 and 95, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 93, 94 and 95, respectively, and the second light chain variable region comprises L2CDR1, L2CDR2 and L2CDR3 comprising amino acid sequences set forth in SEQ ID NOs: 96, 97 and 98, or amino acid sequences having at least 85% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 96, 97 and 98, respectively.

4. The anti-CD47/anti-PD-L1 antibody according to claim 1, wherein: (1) the first heavy chain variable region comprises an amino acid sequence selected from the group consisting of: (b1) amino acid sequences set forth in SEQ ID NO: 22 and SEQ ID NO: 30, (b2) amino acid sequences derived from the amino acid sequences set forth in (b1) by substitution, deletion or addition of one or more amino acids and functionally identical or similar to the amino acid sequences set forth in (b1), and (b3) amino acid sequences having at least 80% sequence identity to the amino acid sequences set forth in (b1); and (2) the first light chain variable region comprises an amino acid sequence selected from the group consisting of: (b4) amino acid sequences set forth in SEQ ID NO: 25 and SEQ ID NO: 33, (b5) amino acid sequences derived from the amino acid sequences set forth in (b4) by substitution, deletion or addition of one or more amino acids and functionally identical or similar to the amino acid sequences set forth in (b4), and (b6) amino acid sequences having at least 80% sequence identity to the amino acid sequences set forth in (b4).

5. The anti-CD47/anti-PD-L1 antibody according to claim 1, comprising an anti-PD-L1 antibody or antigen-binding fragment thereof, wherein: the anti-PD-L1 antibody or antigen-binding fragment thereof comprises a second heavy chain variable region and a second light chain variable region, wherein: (1) the second heavy chain variable region comprises an amino acid sequence selected from the group consisting of: (B1) amino acid sequences set forth in SEQ ID NOs: 99, 100, 101, 102, 110, 111, 112, 113, 114, 119, 120, 121, 122 and 123, (B2) amino acid sequences derived from the amino acid sequences set forth in (B1) by substitution, deletion or addition of one or more amino acids and functionally identical or similar to the amino acid sequences set forth in (B1), and (B3) amino acid sequences having at least 80% sequence identity to the amino acid sequences set forth in (B1); and (2) the second light chain variable region comprises an amino acid sequence selected from the group consisting of: (B4) amino acid sequences set forth in SEQ ID NOs: 103, 104, 105, 106, 115, 116, 117, 118, 124, 125 and 126, (B5) amino acid sequences derived from the amino acid sequences set forth in (B4) by substitution, deletion or addition of one or more amino acids and functionally identical or similar to the amino acid sequences set forth in (B4), and (B6) amino acid sequences having at least 80% sequence identity to the amino acid sequences set forth in (B4).

6. The anti-CD47/anti-PD-L1 antibody according to claim 1, wherein: (1) the first heavy chain variable region comprises an amino acid sequence selected from the group consisting of: (c1) an amino acid sequence set forth in SEQ ID NO: 30, (c2) amino acid sequences derived from the amino acid sequences set forth in (c1) by substitution, deletion or addition of one or more amino acids and functionally identical or similar to the amino acid sequences set forth in (c1), and (c3) amino acid sequences having at least 80% sequence identity to the amino acid sequences set forth in (c1); and (2) the first light chain variable region comprises an amino acid sequence selected from the group consisting of: (c4) an amino acid sequence set forth in SEQ ID NO: 33, (c5) amino acid sequences derived from the amino acid sequences set forth in (c4) by substitution, deletion or addition of one or more amino acids and functionally identical or similar to the amino acid sequences set forth in (c4), and (c6) amino acid sequences having at least 80% sequence identity to the amino acid sequences set forth in (c4).

7. The anti-CD47/anti-PD-L1 antibody according to claim 1, comprising an anti-PD-L1 antibody or antigen-binding fragment thereof, wherein: the anti-PD-L1 antibody or antigen-binding fragment thereof comprises a second heavy chain variable region and a second light chain variable region, wherein: the second heavy chain variable region comprises an amino acid sequence selected from the group consisting of: amino acid sequences set forth in SEQ ID NOs: 110, 111, 112, 113 and 114, amino acid sequences derived from SEQ ID NO: 110, 111, 112, 113 or 114 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 110, 111, 112, 113 or 114, and amino acid sequences comprising H2CDR1, H2CDR2 and H2CDR3 set forth in SEQ ID NOs: 75, 76 and 77 and having at least 85% sequence identity to SEQ ID NO: 110, 111, 112, 113 or 114; and the second light chain variable region comprises an amino acid sequence selected from the group consisting of: amino acid sequences set forth in SEQ ID NOs: 115, 116, 117 and 118, amino acid sequences derived from SEQ ID NO: 115, 116, 117 or 118 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 115, 116, 117 or 118, and amino acid sequences comprising L2CDR1, L2CDR2 and L2CDR3 set forth in SEQ ID NOs: 78, 79 and 80 and having at least 85% sequence identity to SEQ ID NO: 115, 116, 117 or 118; or the second heavy chain variable region comprises an amino acid sequence selected from the group consisting of: amino acid sequences set forth in SEQ ID NOs: 119, 120, 121, 122 and 123, amino acid sequences derived from SEQ ID NO: 119, 120, 121, 122 or 123 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 119, 120, 121, 122 or 123, and amino acid sequences comprising H2CDR1, H2CDR2 and H2CDR3 set forth in SEQ ID NOs: 87, 88 and 89 and having at least 85% sequence identity to SEQ ID NO: 119, 120, 121, 122 or 123; and the second light chain variable region comprises an amino acid sequence selected from the group consisting of: amino acid sequences set forth in SEQ ID NOs: 124, 125 and 126, amino acid sequences derived from SEQ ID NO: 124, 125 or 126 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 124, 125 or 126, and amino acid sequences comprising L2CDR1, L2CDR2 and L2CDR3 set forth in SEQ ID NOs: 90, 91 and 92 and having at least 85% sequence identity to SEQ ID NO: 124, 125 or 126.

8. The anti-CD47/anti-PD-L1 antibody according to claim 1, comprising an anti-PD-L1 antibody or antigen-binding fragment thereof, wherein: the anti-PD-L1 antibody or antigen-binding fragment thereof comprises a second heavy chain variable region and a second light chain variable region, wherein: the second heavy chain variable region comprises an amino acid sequence selected from the group consisting of: an amino acid sequence set forth in SEQ ID NO: 112, amino acid sequences derived from SEQ ID NO: 112 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 112, and amino acid sequences comprising H2CDR1, H2CDR2 and H2CDR3 set forth in SEQ ID NOs: 75, 76 and 77 and having at least 85%, at least 90%, at least 95%, or at least 98% sequence identity to SEQ ID NO: 112; and the second light chain variable region comprises an amino acid sequence selected from the group consisting of: an amino acid sequence set forth in SEQ ID NO: 116, amino acid sequences derived from SEQ ID NO: 116 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 116, and amino acid sequences comprising L2CDR1, L2CDR2 and L2CDR3 set forth in SEQ ID NOs: 78, 79 and 80 and having at least 85%, at least 90%, at least 95%, or at least 98% sequence identity to SEQ ID NO: 116; the second heavy chain variable region comprises an amino acid sequence selected from the group consisting of: an amino acid sequence set forth in SEQ ID NO: 123, amino acid sequences derived from SEQ ID NO: 123 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 123, and amino acid sequences comprising H2CDR1, H2CDR2 and H2CDR3 set forth in SEQ ID NOs: 87, 88 and 89 and having at least 85%, at least 90%, at least 95%, or at least 98% sequence identity to SEQ ID NO: 123; and the second light chain variable region comprises an amino acid sequence selected from the group consisting of: an amino acid sequence set forth in SEQ ID NO: 126, amino acid sequences derived from SEQ ID NO: 126 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 126, and amino acid sequences comprising L2CDR1, L2CDR2 and L2CDR3 set forth in SEQ ID NOs: 90, 91 and 92 and having at least 85%, at least 90%, at least 95%, or at least 98% sequence identity to SEQ ID NO: 126; or the second heavy chain variable region comprises an amino acid sequence selected from the group consisting of: an amino acid sequence set forth in SEQ ID NO: 102, amino acid sequences derived from SEQ ID NO: 102 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 102, and amino acid sequences comprising H2CDR1, H2CDR2 and H2CDR3 set forth in SEQ ID NOs: 93, 94 and 95 and having at least 85%, at least 90%, at least 95%, or at least 98% sequence identity to SEQ ID NO: 102; and the second light chain variable region comprises an amino acid sequence selected from the group consisting of: an amino acid sequence set forth in SEQ ID NO: 106, amino acid sequences derived from SEQ ID NO: 106 by substitution, deletion or addition of one or more amino acids and functionally identical to SEQ ID NO: 106, and amino acid sequences comprising L2CDR1, L2CDR2 and L2CDR3 set forth in SEQ ID NOs: 96, 97 and 98 and having at least 85%, at least 90%, at least 95%, or at least 98% sequence identity to SEQ ID NO: 106.

9. The anti-CD47/anti-PD-L1 antibody according to claim 1, wherein the antibody is a humanized antibody or a fully human antibody.

10. The anti-CD47/anti-PD-L1 antibody according to claim 1, wherein the antibody is a bispecific antibody.

11. An isolated nucleic acid encoding the anti-CD47/anti-PD-L1 antibody according to claim 1.

12. The nucleic acid according to claim 11, wherein (1) the nucleotide sequence encoding the amino acid sequence of the first heavy chain variable region is set forth in SEQ ID NO: 36; (2) the nucleotide sequence encoding the amino acid sequence of the first light chain variable region is set forth in SEQ ID NO: 39; (3) the nucleotide sequence encoding the amino acid sequence of the second heavy chain variable region is set forth in SEQ ID NO: 127 or SEQ ID NO: 128; and (4) the nucleotide sequence encoding the amino acid sequence of the second light chain variable region is set forth in SEQ ID NO: 129 or SEQ ID NO: 130.

13. An expression vector comprising the nucleic acid according to claim 11.

14. A host cell transformed with the expression vector according to claim 13, wherein the host cell is selected from the group consisting of prokaryotic cells and eukaryotic cells.

15. A method for producing the anti-CD47/anti-PD-L1 antibody according to claim 1, comprising expressing the antibody in the host cell according to claim 14, and isolating the antibody from the host cell.

16. A pharmaceutical composition comprising the anti-CD47/anti-PD-L1 antibody according to claim 1 and a pharmaceutically acceptable carrier.

17. A method for treating a cancer comprising administering the anti-CD47/anti-PD-L1 antibody according to claim 1 to the subject in need thereof.

18. The method according to claim 17, wherein the cancer is selected from the group consisting of hematological tumor, lymphoma, breast cancer, lung cancer, gastric cancer, intestinal cancer, esophageal cancer, ovarian cancer, cervical cancer, kidney cancer, bladder cancer, pancreatic cancer, glioma and melanoma.

19. The host cell according to claim 14, wherein the cell is a mammalian cell.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0223] FIG. 1 shows the results of the binding activity assay (ELISA) of the humanized anti-CD47 antibodies to monkey CD47.

[0224] FIG. 2 shows the results of the binding activity assay (ELISA) of the humanized anti-CD47 antibodies to human CD47.

[0225] FIG. 3 shows the results of the binding activity assay (ELISA) of the humanized anti-CD47 antibodies to CD47 on cell surface.

[0226] FIG. 4 shows the results of the hemagglutination test, wherein RBC represents the positive control and PBS represents the blank control.

[0227] FIG. 5 shows the results of the blocking activity of the humanized anti-CD47 antibodies detected by FACS.

[0228] FIG. 6 shows the results of the anti-tumor test of the humanized anti-CD47 antibody Hu34-39-PE in human gastric cancer NUGC-4 xenograft model.

[0229] FIG. 7 shows the results of the anti-tumor test of the humanized anti-CD47 antibody Hu26T-31-PE in human gastric cancer NUGC-4 xenograft model.

[0230] FIG. 8 shows a schematic diagram of the structure of bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs).

[0231] FIG. 9 shows a schematic diagram of the sequence of ScFabHuPL7-21Ks.

[0232] FIG. 10 shows a schematic diagram of the sequence of ScFabHu34-39Hs.

[0233] FIG. 11 shows the results of the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) binding to PD-L1 (A), competing with PD-1 (B), binding to CD47 (C), competing with SIRPα (D) and competing with CD80 (E).

[0234] FIG. 12 shows the results of the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) binding to PD-L1 on cell surface (A), binding to CD47 on cell surface (B), blocking the binding of PD-1 to PD-L1 (C) and blocking the binding of SIRPα to CD47 (D) at the cellular level.

[0235] FIG. 13 shows the results of the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) binding to both CD47 and PD-L1 on Raji-hPD-L1 cells (A, B) and blocking both CD47 and PD-L1 (B).

[0236] FIG. 14 shows the results of the bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) inhibiting the growth of transplanted tumors in mice.

DETAILED DESCRIPTION

[0237] The following representative examples are used to illustrate the present disclosure better, rather than to limit the scope of protection of the present disclosure. The experimental methods without conditions indicated in the following examples are usually carried out according to conventional conditions, such as the antibody technology experiment manual and molecular cloning manual of Cold Spring Harbor, or according to the conditions recommended by the raw material or commodity manufacturers. The materials and reagents used in the examples are all commercially available unless otherwise specified.

Example 1 Preparation of CD47 Antigen Protein and Anti-CD47 Positive Control Antibody

1. Construction of Expression Vector of Antigen and Positive Control Antibody

(1) Construction of Expression Vector of Antigen

[0238] A gene fragment encoding the full length of CD47 protein was synthesized, and the amino acid sequence is shown in SEQ ID NO: 41. This fragment was cloned into the eukaryotic expression plasmid pTargeT to generate the CD47 expression plasmid pTargeT-CD47.

[0239] The amino acid sequence of the extracellular region of the human CD47 protein was fused with the amino acid sequence of hIgG1-Fc or his-tag, and the designed amino acid sequences are shown in SEQ ID NO: 42 and SEQ ID NO: 43, respectively. After codon optimization of the above sequences encoding amino acids, the tagged extracellular region of CD47 protein coding fragments, CD47-hFc and CD47-his, were synthesized and cloned into the eukaryotic expression plasmid pHR respectively to generate expression plasmids pHR-CD47-hFc and pHR-CD47-his.

[0240] The amino acid sequence of the extracellular region of the human CD47 protein was fused with the amino acid sequence of mIgG1-Fc, and the designed amino acid sequence is shown in SEQ ID NO: 44. After codon optimization of the sequence, a complete expression plasmid pcDNA3.1(+)-TPA-CD47-mIgG1-Fc was constructed.

[0241] The sequence of SIRPα is shown in SEQ ID NO: 45. After codon optimization of the sequence, a complete expression plasmid pcDNA3.1(+)-SIRPα-myc-His was constructed.

(2) Construction of Expression Vector of Positive Control Antibody

[0242] The antibody AB6.12-IgG4P (abbreviated as AB06.12-4P hereinafter) disclosed in the patent application WO2013/119714 was used as a positive control antibody. The amino acid sequences of AB06.12-4P are as follows:

[0243] heavy chain amino acid sequence of AB06.12-4P is shown in SEQ ID NO: 46; and

[0244] light chain amino acid sequence of AB06.12-4P is shown in SEQ ID NO: 47.

[0245] The amino acid sequence corresponding to the above antibody sequences were artificially optimized to generate the heavy chain and light chain expression plasmids pcDNA3.1(+)-SHC025-hG4 and pcDNA3.1(+)-SHC025-hk of the positive control antibody AB06.12-4P. The heavy chain fragment was cloned into the eukaryotic expression plasmid pHR containing the light chain constant region of IgG4 to obtain the heavy chain eukaryotic expression plasmid pHR-SHC025-hG4-4PE and the light chain expression plasmid pcDNA3.1(+)-SHC025-hk of AB06.12-4P.

2. Expression and Purification of Antigen Protein and Positive Control Antibody

(1) Construction of Stable Transgenic Cell Line of Antigen Protein

[0246] CHO-K1 cells (Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences) were electrotransfected with eukaryotic expression plasmid pTargeT-CD47 under a square pulse of 15 msec at a voltage of 160V and then cultured in an incubator at 37° C. with 5% CO.sub.2. After 24 h, the cells were subjected to selection with culture medium containing 500 μg/ml G418. After 16 days, the positive rate of cell pool was detected by FACS. The cells transfected with plasmid were plated (1×10.sup.6 cells/ml of cell density, 100 μl/well), and incubated with PE mouse anti-human CD47 antibody (BD, 556046). Flow cytometer (BD, FACSJazz) was used to read the mean value at a wavelength of 585 nm, and data analysis was performed using GraphPad. The positive cell lines were subcloned, and a CHO-K1 cell line was selected, which expressed CD47 at a high level and was named CHO-K1-E5.

(2) Expression of Tagged Antigen Protein and Positive Control Antibody

[0247] 293F cells were inoculated into a 1 L cell culture flask with a density of 0.5×10.sup.6 cells/ml. Fresh and pre-warmed FreeStyle 293 expression medium was added to make the total volume 250 mL. The cells were cultured in a humidified CO.sub.2 incubator at 37° C. with 8% CO.sub.2 overnight. 500 μl of 1 mg/ml PEI solution was added to 8.5 mL FreeStyle 293 expression medium and mixed well. 250 μg of plasmid to be transfected was added to 8.5 ml FreeStyle 293 expression medium and mixed well, wherein, the plasmids encoding tagged antigen, pHR-CD47-hFc, pHR-CD47-his, pcDNA3.1(+)-TPA-CD47-mIgG1-Fc and pcDNA3.1(+)-SIRPα-myc-His, were transfected respectively. The heavy chain plasmid pHR-SHC025-hG4-4PE and the light chain plasmid pcDNA3.1(+)-SHC025-hk of positive control antibody AB06.12-4P were co-transfected. The FreeStyle 293 expression medium containing PEI was added to the expression medium containing plasmids and mixed well, then the mixture was added to the cells, and the cells were cultured in a humidified CO.sub.2 incubator at 37° C. with 8% CO.sub.2. The cells were fed on the 1st and 3rd day after cell transfection with 2.5 ml of 200 mM glutamine and 5 ml of 180 g/L glucose per flask. The cell culture supernatant was collected when the cell viability dropped to 65%-75%. The cell culture was centrifuged at 1,500 rpm for 5 min to collect the supernatant. The supernatants were centrifuged at 8,000 rpm for 20 min to collect the supernatant again.

(3) Affinity Chromatography Purification

[0248] Different affinity chromatography columns were used to perform purification by using AKTA machine (GE, AKTA pure-150) according to the properties of the proteins (see Table 1 for affinity chromatography columns suitable for different proteins). The specific purification steps are as follows.

TABLE-US-00001 TABLE 1 Affinity chromatography columns suitable for different proteins Protein Column Brand Model Murine monoclonal antibody Protein G Bestchrom Ezfast (hybridoma)/CD47-mFc prepacked Protein column G 4FF CD47-his/SIRPα-myc-His NI prepacked GE His Trap, column HP 5 ml CD47-hFc/chimeric antibody, Protein A GE Hi Trap, humanized antibody prepacked Mabselect column SuRe 5 ml Protein A GE Mabselect self-packed LX 18 ml column

[0249] Cleaning

[0250] The equipment and pipelines were cleaned with ultrapure water for 2 min with a flow rate of 10 mL/min, and then the chromatography system was cleaned with 0.1M NaOH.

[0251] Column Connection

[0252] The chromatography column was connected to the chromatography equipment and rinsed with ultrapure water for 5 min, and then the chromatography system was rinsed with 0.1M NaOH for 30 min with a retention time of 5 min.

[0253] Equilibration

[0254] Five CVs (column volume) of 20 mM PB+0.15M NaCl, pH 7.2 was used to equilibrate the column.

[0255] Sample Loading

[0256] The supernatant from cell culture was loaded to the column with a retention time of 5 min.

[0257] Post-Equilibration

[0258] Five CVs 20 mM PB+0.15M NaCl, pH 7.2 was used to equilibrate.

[0259] Elution

[0260] Elution was performed with 50 mM acetic acid (pH=3.4), for a retention time of 5 min. Collection started when UV280 reached about 50 mAu, and stopped when UV280 dropped to about 50 mAu. The sample was adjusted to the pH of 7.0 with 1M Tris-HCl (pH 9.0).

[0261] Re-Equilibration

[0262] Three CVs of 20 mM PB+0.15M NaCl, pH 7.2, was used to equilibrate with a retention time of 5 min.

[0263] On-Column Cleaning

[0264] Cleaning was performed with 0.1M NaOH for 30 min with a retention time of 5 min.

[0265] Cleaning and Preservation

[0266] Cleaning was performed with purified water for 10 minutes, and then 2 CVs of 20% ethanol.

Example 2 Preparation of Anti-CD47 Monoclonal Antibodies

1. Preparation of Hybridomas

(1) Animal Immunization

[0267] The experimental SJL mice were immunized with different tagged CD47 proteins together with adjuvants. 50 μg of antigen was used for the first shoot, and 25m of antigen was used for the subsequent immunization.

[0268] The immune adjuvants used in the experiments may be QuickAntibody-Mouse5W (Beijing Biodragon immunotech. Co., Ltd.), TiterMax (Sigma), CpG (GenScript Biotechnology Co., Ltd.), or Alum (thermo) adjuvant. Different tagged CD47 protein samples were added dropwise to the adjuvant solution with vortex to mix thoroughly. The dosages of the adjuvant were referred to the instructions. After the mixture was mixed well and formed a water-in-oil emulsion, the SJL mice were immunized.

[0269] Cell lines expressing high level of CD47, such as CCRF-CEM and CHO-K1-E5, were also used to immunize mice to produce antibodies. The cultured human acute lymphoblastic leukemia cells (CCRF-CEM) and CHO-K1-E5 (positive cell line obtained in Example 1) were treated with trypsin and then centrifuged at 1,000 rpm for 5 min. The supernatant was discarded and the cell pellets were resuspended in PBS. Part of the cell sample was taken out for cell counting, and the remaining cell sample was centrifuged at 1,000 rpm for 5 min. The supernatant was discarded and the cell pellet was resuspended in PBS. An appropriate amount of PBS was added to obtain a cell suspension of 1×10.sup.8 cells/ml. Each mouse in the experimental group was immunized with 1×10.sup.7 cells.

[0270] The immunization protocol is shown in Table 2.

TABLE-US-00002 TABLE 2 Mouse Immunization Protocol Group Antigen Adjuvant Route* 1 PBS None 2 CD47-his/CD47-mFc Quick Antibody-Mouse5W i.m. 3 CD47-his/CD47-mFc Titer Max/CpG/Alum s.c./i.m. 4 CD47-mFc Quick Antibody-Mouse5W i.m. 5 CD47-mFc Titer Max/CpG/Alum s.c./i.m. 6 CCRF-CEM/CHO- None i.p. K1-E5 *i.m.: intramuscular injection; s.c.: subcutaneous injection; i.p.: intraperitoneal injection.

(2) Hybridoma Fusion

[0271] Acquisition and preparation of spleen cells. The mice after booster immunization were sacrificed and soaked in 75% alcohol. The spleen was dissected out, ground with a grinding rod, and filtered through a cell strainer to prepare a single cell suspension. The spleen cell suspension was centrifuged at 2,000 rpm for 5 min, and the supernatant was discarded. 2 mL red blood cell lysate was added to lyse red blood cells at room temperature for 2 min and PBS was added to reach 20 mL. After centrifugation at 1,500 rpm for 7 min, the supernatant was discarded. Viable cells were counted after resuspension. The Sp2/0 cells in the culture flask were collected and after centrifuged at 1,000 rpm for 5 min, the supernatant was discarded. Viable cells were counted after resuspension. The spleen cells were mixed with Sp2/0 cells at a ratio of 1:1 and subjected to centrifugation at 1,500 rpm for 7 min, the supernatant was discarded. The cells were resuspended in 20 mL electroporation buffer. After centrifugation at 1,500 rpm for 7 min, the supernatant was discarded and the step was repeated once. The cells were resuspended with an appropriate amount of electroporation buffer to ensure the cell concentration of about 2×10.sup.7 cells/mL. The cell suspension was added to a 9 mL electroporation tank for fusion. After fusion, the cell suspension was transferred to 15 mL RPMI 1640 complete medium containing 20% FBS and then left at room temperature for 20 min. The fused cells were resuspended with RPMI 1640 medium containing 1×HAT, 1×BIOMYC3, and 20% FBS. The cell suspension was added to several 96-well cell culture plates at 100 μl/well to ensure that the cell volume per well was about 4×10.sup.4 cells/well, and the plates was placed in a 37° C. cell incubator. After 5 days, additional 100 μL of RPMI 1640 complete medium containing 20% FBS, 1×HAT, and 1×BIOMYC-3 was added to each well.

(3) Screening of Hybridoma and Subcloning

[0272] After one week of fusion, the supernatants of culture were collected and used for screening the hybridoma supernatants that can bind to CD47-his protein or CD47 on cell surface by ELISA. CD47-his was used to screen for antibodies against CD47 instead of hFc and mFc. The ability of the hybridoma supernatant to block the CD47-SIRPα interaction was analyzed by ELISA. SIRPα-myc-his was coated on ELISA plates. The mixture of recombinant humanized CD47-hFc and hybridoma supernatant was added and incubated for 2 h. HRP-labeled anti-human IgG Fc specific antibody (Jackson Immuno Research) was added and incubated for 1 h. Microplate reader was used to detect absorbance at 450 nm. The hybridoma parent clones showing binding and blocking activities in the screening experiments were expanded. The binding and blocking activities were retested, and the hybridoma positive clones with binding and blocking activities were obtained by screening again.

[0273] The positive cell lines were subcloned by the limiting dilution method. After one week of culture, the binding activity to CD47 and the activity of blocking the CD47-SIRPα interaction of the supernatants were detected by ELISA. Three cell lines that showed positive results in the above two tests were obtained, respectively named as SHC025-26, SHC025-34 and SHC025-58.

2. Identification of Antibody Subtypes

[0274] The antibody subtypes were identified according to the instructions of the mouse antibody subtype identification kit “SBA Clonotyping Systerm-057BL/6-HRP” (SouthernBiotech, Cat. No. 5300-05B). The results are shown in Table 3.

TABLE-US-00003 TABLE 3 Identification results of antibody subtypes Antibody Antibody subtype SHC025-26 IgG1/k SHC025-34 IgG2c/k SHC025-58 IgG2b/k

3. Preparation of Monoclonal Antibodies

[0275] According to the activity analysis results of the supernatants from cell culture, the parent clones of monoclonal antibodies SHC025-26, SHC025-34, and SHC025-58 were identified and expanded. The culture medium was 1640 medium containing 10% fetal bovine serum, 1×NAEE, 1× sodium pyruvate, and 1% penicillin-streptomycin double antibiotics. When the cell confluence was >80%, the cells were subcultured and expanded. 50 ml of the supernatant was collected and the antibody was purified. The obtained antibody was subjected to SDS-PAGE gel electrophoresis and showed a good purity.

4. Sequencing of Monoclonal Antibodies

[0276] The subcloned positive hybridomas were expanded, and an appropriate amount of cells was used for total RNA extraction according to the instructions of RNeasy Plus Mini Kit (Qiagen, 74134). The first strand of cDNA was synthesized using Prime Script 1st strand cDNA Synthesis Kit (Takara, 6110A).

[0277] Specific primers were designed according to the variable region of the mouse antibody subtype, and 5′ end of the primers contained the homologous arm sequence for homologous recombination with the eukaryotic expression vector. PCR amplification for the variable region of antibodies was performed using cDNA as a template to obtain the gene fragments of the light chain variable region and heavy chain variable region of the mouse antibody respectively. The design of primers refers to references: 1. Anke Krebber, Susanne Bornhauser, Jorg Burmester etal. Reliable cloning of functional antibody variable domains from hybridomas and spleen cell repertoires employing a reengineered phage display system. Journal of Immunological Methods, 1997, 201: 35-55; 2. Simon KorenMiha KosmaĉAnja Colja Venturini etal. Antibody variable-region sequencing as a method for hybridoma cell-line authentication, 2008, 78: 1071-1078. DNA sequencing was performed and the results are shown in Table 4.

TABLE-US-00004 TABLE 4 Sequence table of anti-CD47 murine-derived monoclonal antibody Amino acid sequence of the Amino acid sequence of the Antibody heavy chain variable region light chain variable region SHC025-26 SEQ ID NO: 21 SEQ ID NO: 24 SHC025-34 SEQ ID NO: 22 SEQ ID NO: 25 SHC025-58 SEQ ID NO: 23 SEQ ID NO: 26

[0278] For the antibody SHC025-34, the CDR sequences of VH: the sequences of CDR1, CDR2 and CDR3 are set forth in SEQ ID NOs: 4, 5 and 6, respectively; the CDR sequences of VL: the sequences of CDR1, CDR2 and CDR3 are set forth in SEQ ID NOs: 15, 16 and 17, respectively.

Example 3 Construction of Anti-CD47 Chimeric Antibodies

[0279] The purified gene fragments of the light chain and the heavy chain variable regions of the mouse antibody (see Example 1 for the purification steps) were respectively co-transformed into Escherichia coli DH5a competent cells with the linearized eukaryotic expression plasmid containing the light chain constant region or the heavy chain constant region of the human antibody. The mixture was spread evenly on the surface of the agar plates containing the corresponding antibiotics. The agar plates were incubated in a 37° C. constant temperature incubator overnight, and then several single colonies were picked out for DNA sequencing. The chimeric antibodies with correct sequences were named SHC025-26CHI, SHC025-34CHI, and SHC025-58CHI.

[0280] The positive clones with correct sequences were inoculated in 2×YT liquid medium containing the corresponding antibiotics and cultured at 37° C. for more than 12 hours with shaking. The bacterial cells were collected for plasmid extraction to obtain the plasmids expressing the light chain and the heavy chain of the chimeric antibody. The concentration and purity of the plasmids were detected by a nucleic acid quantitative analyzer.

[0281] The plasmids for chimeric antibodies were transfected into HEK293E cells, and the antibodies were expressed and purified. The purity, activity and affinity were tested and analyzed.

[0282] By sequencing, it was found that there was one cysteine in the heavy chain CDR position 118 of SHC025-26 and one cysteine in the light chain CDR position 56 of SHC025-58. During the antibody expression, the cysteine in CDR region would randomly pair with the other cysteine on the antibody molecule to form a disulfide bridge, which thereby would greatly affect the purity of the antibody. To solve this problem, the amino acid sequences of SHC025-26CHI and SHC025-58CHI were modified as follows: C118 of SHC025-26CHI heavy chain was mutated to T and the obtained antibody was named SHC025-26CHI-T; C56 of SHC025-58CHI light chain was mutated to A and the obtained antibody was named SHC025-58CHI-A. The mutant sequences were constructed by site-directed mutagenesis. The results of the chimeric antibody sequencing are shown in Table 5.

TABLE-US-00005 TABLE 5 Sequences of anti-CD47 chimeric antibodies Amino Chimeric antibody acid sequence of VH Amino acid sequence of VL SHC025-26CHI SEQ ID NO: 21 SEQ ID NO: 24 SHC025-26CHI-T SEQ ID NO: 27 SEQ ID NO: 24 SHC025-34CHI SEQ ID NO: 22 SEQ ID NO: 25 SHC025-58CHI SEQ ID NO: 23 SEQ ID NO: 26 SHC025-58CHI-A SEQ ID NO: 23 SEQ ID NO: 28

[0283] For the antibody SHC025-34CHL the sequences of CDR1, CDR2 and CDR3 of VH are set forth in SEQ ID NOs: 4, 5 and 6, respectively; the sequences of CDR1, CDR2 and CDR3 of VL are set forth in SEQ ID NOs: 15, 16 and 17, respectively.

Example 4 Construction and Production of Humanized Anti-CD47 Antibodies

[0284] Based on the results of activity analysis and affinity KD value of chimeric antibodies, SHC025-34CHI, SHC025-58CHI-A, and SHC025-26CHI-T were modified to humanized antibodies.

[0285] To construct the humanized antibodies, the variable regions of SHC025-34CHI, SHC025-58CHI-A, and SHC025-26CHI-T antibody were compared with the mouse antibody sequences in the ImMunoGeneTics (IMGT) to determine their murine-derived germlines. After homology comparison, it was found that the FR region sequences of the heavy chain variable region of SHC025-34CHI, SHC025-58CHI-A, and SHC025-26CHI-T antibody were the most similar to the human antibody germline genes IGHV1-8*01, IGHV3-21*04, and IGHV1-2*02 respectively; the FR sequences of the light chain variable region of the antibodies were the most similar to the human antibody germline genes IGKV3-11*01, IGKV1-5*01 and IGKV4-1*01 respectively. With the framework region sequence FR1-FR3 of SHC025-34CHI/SHC025-58CHI-A antibody as templates, full human framework regions with similar 3D structure but low immunogenicity were screened in the human framework region library to replace FR1-FR3 sequence of SHC025-34CHI/SHC025-58CHI-A. The full-length sequences of the heavy/light chain were 3D modeled and compared structurally with the heavy/light chain sequences of the original antibodies. Considering the antigenicity and 3D structural similarity, 6 humanized heavy chain variable regions (see SEQ ID NOs: 48, 49, 50, 51, 52, and 53) and 4 humanized light chain variable regions (see SEQ ID NOs: 54, 55, 56, and 57) of SHC025-34CHI, and 6 humanized heavy chain variable regions (see SEQ ID NOs: 58, 59, 60, 61, 62, and 63) and 5 humanized light chain variable regions (see SEQ ID NOs: 64, 65, 66, 67, and 68) of SHC025-58CHI-A were ultimately selected for further optimization. More than 95% of non-CDR regions of SHC025-34CHI or SHC025-58CHI-A antibody were humanized. The variable region sequences of the heavy and light chain of SHC025-26CHI-T were used to perform structural alignment analysis in Protein Data Bank. The FR1-FR3 sequences with the highest similarity were selected to replace the murine-derived sequences, and the amino acid sites that displayed key role in structure stabilization of the antibody in the structural simulation were mutated back to murine-derived amino acid residues. Finally, 4 humanized heavy chain variable regions (see SEQ ID NOs: 69, 70, 71, and 72) and 2 humanized light chain variable regions (see SEQ ID NOs: 73 and 74) of SHC025-26CHI-T were obtained.

[0286] The amino acid sequences of the light chain and heavy chain variable region of humanized antibody obtained above were reversely transcribed into their corresponding nucleotide sequences, and oligonucleotide fragments containing complementary sequences between adjacent fragments were generated. The oligonucleotide fragments were annealed and assembled by Overlap PCR. Then nucleotide fragments of the entire light chain and heavy chain variable regions were amplified using specific primers (5′ end contained the homologous arm sequence for homologous recombination with the eukaryotic expression vector). The purified nucleotide fragments of the light chain variable region were co-transformed into Escherichia coli DH5a competent cells with the linearized eukaryotic expression plasmid containing light chain constant region of IgG4. The purified nucleotide fragments of the heavy chain variable region were co-transformed into Escherichia coli DH5a competent cells with the eukaryotic expression plasmid containing heavy chain constant region of IgG4 containing S228P/L235E mutation. The competent cells with the transformed plasmid were spread evenly on the surface of the agar plates containing the corresponding antibiotics. The agar plates were incubated in a 37° C. constant temperature incubator overnight, and then several single colonies were picked out for DNA sequencing.

[0287] The positive clones with correct sequences were inoculated in 2×YT liquid medium containing the corresponding antibiotics and cultured at 37° C. with shaking for more than 12 hours. The bacterial cells were collected for plasmid extraction to obtain the expression plasmids for the light chain and the heavy chain of humanized antibodies. The concentration and purity of the plasmids were detected by a nucleic acid quantitative analyzer.

[0288] The plasmids were transfected into HEK293E cells, and a large number of antibodies were expressed and purified. The purity, activity and affinity were tested and analyzed.

[0289] The humanized antibodies with good purity, activity and affinity were selected, and named as Hu26T-31-PE, Hu34-39-PE, and Hu58A-14-PE. The sequences are shown in Table 6. The sources of humanized antibodies are shown in Table 7.

TABLE-US-00006 TABLE 6 Sequence table of humanized anti-CD47 antibodies Humanized Amino acid sequence Nucleotide sequence antibody VH VL VH VL Hu26T-31- SEQ ID NO: 29 SEQ ID SEQ ID NO: 35 SEQ ID PE NO: 32 NO: 38 Hu34-39-PE SEQ ID NO: 30 SEQ ID SEQ ID NO: 36 SEQ ID NO: 33 NO: 39 Hu58A-14- SEQ ID NO: 31 SEQ ID SEQ ID NO: 37 SEQ ID PE NO: 34 NO: 40

[0290] For the antibody Hu34-39-PE, the sequences of CDR1, CDR2 and CDR3 of VH are set forth in SEQ ID NOs: 4, 5 and 6, respectively; the sequences of CDR1, CDR2 and CDR3 of VL are set forth in SEQ ID NOs: 15, 16 and 17, respectively.

TABLE-US-00007 TABLE 7 Information of humanized sequences Source of humanized sequence Recombinant humanized sequence Heavy chain X62106 Homsap IGHV1-2*02 F SEQ ID NO: 29 X92343 Homsap IGHV1-46*01 F SEQ ID NO: 30 HM855688 Homsap SEQ ID NO: 31 IGHV3-21*04 F Light chain X97473 Homsap IGLV3-9*01 F SEQ ID NO: 32 X71966 Homsap IGLV3-21*01 F SEQ ID NO: 33 Z73648 Homsap IGLV4-69*01 F SEQ ID NO: 34

Example 5 Binding Activity Assay of Anti-CD47 Antibodies to Monkey CD47 (by ELISA)

[0291] The binding activity of antibodies was analyzed by protein based ELISA. Cynomolgus CD47-His (0.1 μg/well, ACRO Biosystems, Cat. No. CD7-052H1-50m) was coated on 96-well ELISA plates. The anti-CD47 antibodies of the present disclosure was used as the primary antibody and added to the ELISA plates in 5-fold gradient dilution with a total of 8 concentrations: 2000 ng/mL, 400 ng/mL, 80 ng/mL, 16 ng/mL, 3.2 ng/mL, 0.64 ng/mL, 0.128 ng/mL, and 0 ng/mL respectively. The plates were incubated at 37° C. for 1.5h. AB06.12-4P was used as the positive control antibody and Anti-Human IgG HRP (Jackson, 109-035-003, 1:10000) was used as the secondary antibody. Color developing solution TMB (3,3′,5,5′-tetramethylbenzidine) was added to the plate, and microplate reader (Thermo, Multiskan FC) was used to read OD450 value after termination of the reaction. EC.sub.50 was generated by GraphPad. The result is shown in FIG. 1.

[0292] The experimental results show that humanized anti-CD47 antibodies Hu26T-31-PE, Hu34-39-PE and Hu58A-14-PE of the present disclosure can bind to cynomolgus CD47, and the binding ability is equivalent to that of the positive control antibody AB06.12-4P.

Example 6 Binding Activity Assay of Anti-CD47 Antibodies to Human CD47 (by ELISA)

[0293] The binding activity of antibodies was analyzed by ELISA. Human CD47-His protein (0.1 μg/well, prepared in Examples 1 and 2) was coated on 96-well ELISA plates, and the ELISA plates were incubated at 37° C. for 2 h. After washing 3 times with 1×PBST, the ELISA plates were blocked with 5% non-fat milk at 4° C. overnight. The plates were washed 3 times with 1×PBST. The anti-CD47 antibodies of the present disclosure were used as the primary antibody and added to the ELISA plates at 5-fold gradient dilution with a total of 8 concentrations: 2000 ng/mL, 400 ng/mL, 80 ng/mL, 16 ng/mL, 3.2 ng/mL, 0.64 ng/mL, 0.128 ng/mL, and 0 ng/mL respectively. The plates were incubated at 37° C. for 1.5h. The plates were washed 3 times with 1×PBST. AB06.12-4P was used as the positive control antibody and Anti-Human IgG HRP (Jackson, 109-035-003, 1:10000) was used as the secondary antibody. Then the plates were incubated at 37° C. for 40 min. Color developing solution TMB was added after the plates were washed 5 times with 1×PBST, and microplate reader (Thermo, Multiskan FC) was used to read OD450 value after termination of the reaction. EC.sub.50 was generated by GraphPad. The result is shown in FIG. 2.

[0294] The experimental results show that humanized anti-CD47 antibodies Hu26T-31-PE, Hu34-39-PE and Hu58A-14-PE of the present disclosure can bind to human CD47, and the binding ability is equivalent to that of the positive control antibody AB06.12-4P.

Example 7 Binding Activity Assay of Anti-CD47 Antibodies to CD47 on Cell Surface (by ELISA)

[0295] The binding activity of antibodies was analyzed by cell based ELISA. CHO-K1-E5 cells were plated at 1×10.sup.5 cells per well and cultured overnight at 37° C. and 5% CO.sub.2. On the second day, the cells were fixed with 4% paraformaldehyde, then blocked with non-fat milk for 1 h, and later washed gently with 1×PBS. The anti-CD47 antibody of the present disclosure was used as the primary antibody and added to the plates in 5-fold gradient dilution with a total of 8 concentrations: 2000 ng/mL, 400 ng/mL, 80 ng/mL, 16 ng/mL, 3.2 ng/mL, 0.64 ng/mL, 0.128 ng/mL, and 0 ng/mL respectively. The plates were incubated at 37° C. for 1.5h. AB06.12-4P was used as the positive control antibody and Anti-Human IgG HRP (Jackson, 109-035-003, 1:10,000) was used as the secondary antibody. Color developing solution TMB was added, and microplate reader (Thermo, Multiskan FC) was used to read OD450 value after termination. EC.sub.50 was generated by GraphPad. The result is shown in FIG. 3.

[0296] The experimental results show that humanized anti-CD47 antibodies Hu26T-31-PE, Hu34-39-PE and Hu58A-14-PE of the present disclosure can bind to CD47 on cell surface, and the binding ability is equivalent to that of the positive control antibody AB06.12-4P.

Example 8 Affinity Assay of Anti-CD47 Antibodies to Human CD47 Protein

[0297] The affinity of the humanized anti-CD47 antibodies prepared in Examples 1 and 2 to the antigen CD47(19-136)-hFC was determined by Fortebio Octet. First, the antigen CD47(19-136)-hFc was labeled with biotin, and then put into 10 kD cut-off ultrafiltration tube with PBS for desalting by centrifugation. This step was repeated 3-4 times. The actual concentration of the biotin-labeled antigen (CD47-hFc-Biotin) was determined by Nanodrop machine. CD47-hFc-Biotin was diluted with SD buffer (0.02% Tween20+0.1% BSA solution) to a concentration of 5m/ml. The humanized anti-CD47 antibody was diluted with SD buffer in 4-fold gradient dilution to make the concentrations of 10 μg/ml, 2.5m/ml, 0.625m/ml and 0 μg/ml. SA sensor was used to solidify the antigen, and affinity assay was performed according to the manual of fortebio Octet RED96. The specific parameters and experimental results are shown in Table 8.

TABLE-US-00008 TABLE 8 Affinity assay of antibodies to human CD47 protein Antibody KD (M) kon (l/Ms) kdis (1/s) Hu26T-31-PE 5.87E−11** 1.04E+06 6.13E−05 Hu34-39-PE 1.13E−10 6.69E+06 7.58E−04 Hu58A-14-PE 9.46E−11 1.78E+06 1.68E−04 AB06.12-4P 1.01E−10 4.80E+06 4.85E−04

[0298] The experimental results show that the humanized anti-CD47 antibody Hu26T-31-PE has significantly higher affinity for binding to human CD47 protein than the positive control antibody.

Example 9 Hemagglutination Test of Anti-CD47 Antibodies

[0299] 5 mL blood sample was added to 40 mL PBS. The mixture was centrifuged at 2000 rpm for 5 min and the supernatant was discarded. The cell pellet was washed three times with PBS and then resuspended in PBS. According to the hematocrit, a 2% red blood cell suspension was prepared. The initial concentration of the antibody to be analyzed was 1-20 μM at a 2-fold dilution, a total of 24 concentration gradients. In round-bottom 96-well plates, 50 μL of the above-mentioned antibodies of different concentrations was added, and then 50 ul of the 2% red blood cell suspension was added. The mixture was mixed well, placed at room temperature and monitored for agglutination after 2h. Rabbit polyclonal RBC antibody (Rockland, 109-4139) was used as a positive control for hemagglutination and the results are shown in FIG. 4. As shown in FIG. 4, the concentration of the antibodies (rabbit polyclonal RBC antibody, AB06.12-4P antibody and the test antibodies of the present disclosure) from left to right in the 96-well plates were diluted in a 2-fold gradient starting from 20 uM. RBC indicated the positive control group, in which rabbit polyclonal RBC antibody caused significant hemagglutination), and PBS indicated the blank control group. If there is no cell agglutination, the red blood cells will fall to the bottom of the well as a small dot with smooth edge. Dot with slightly vague edge indicates agglutination of a small amount of red blood cells. If the red blood cells form flaky shape and cover the whole bottom of the well, this indicates agglutination of most of the red blood cells.

[0300] It has been reported that the anti-CD47 antibody Hu5F9-G4 disclosed in the patent application WO2011/143624 can cause significant agglutination of red blood cells within the same concentration range, which is a common undesirable property of anti-CD47 antibodies. However, under the same conditions, Hu26T-31-PE, Hu34-39-PE, and Hu58A-14-PE of the present disclosure did not cause hemagglutination as shown in the experimental results, indicating that the antibodies of the present disclosure are significantly superior to the Hu5F9-G4 antibody in this respect.

Example 10 Blocking Activity Assay of Anti-CD47 Antibodies Against CD47

[0301] The ability of the anti-CD47 antibody provided by the present disclosure to block SIRPα from binding to CD47 on cell surface was detected by FACS. CHO-K1-E5 cells positive for CD47 were used as a CD47 provider. In the presence of a serially diluted anti-CD47 antibody, the binding of CD47 to SIRPα was monitored. PE

[0302] Streptavidin (Biolegend, 405203, 1:200) was used as the secondary antibody to monitor the changes of SIRPα-Biotin, and AB06.12-4P was used as a positive control which blocked SIRPα from binding to CD47 on cell surface. Flow cytometer (BD, FACSJazz) was used to read the mean value at a wavelength of 585 nm and IC.sub.50 was generated by GraphPad. The result is shown in FIG. 5.

[0303] The experimental results show that the blocking activity ranking from high to low is: Hu26T-31-PE≥Hu34-39-PE≥AB06.12-P>Hu58A-14-PE.

Example 11 Anti-Tumor Test of Anti-CD47 Antibodies in Human Gastric Cancer NUGC-4 Xenograft Model

1. Experimental Materials

(1) Experimental Cells and Animals

[0304] NUGC-4 human gastric cancer cells were purchased from American Type Culture Collection (ATCC).

[0305] NOD-Scid mice, female, 5-8 weeks old, weighing 18-20 grams, were purchased from Shanghai Lingchang Biotechnology Co., Ltd.

(2) Test Samples and Controls

[0306] The reference antibody isotype IgG4 (Cat. No. AB170091) was purchased from Crown Bioscience Co., Ltd. and used as a negative control.

[0307] Before the test, the humanized anti-CD47 antibody of the present disclosure was prepared at two concentrations of 0.6 mg/mL and 0.3 mg/mL in PBS, and Isotype IgG4 and AB 06.12-4P were prepared at 0.6 mg/mL.

(3) Experimental Methods

[0308] NUGC-4 human gastric cancer cells were cultured with RMPI1640 medium containing 10% fetal bovine serum, 100 U/mL penicillin and 100 μg/mL streptomycin in a 37° C., 5% CO.sub.2 incubator. The cells were digested, treated with 2 mL 1×EDTA solution and passaged once a week. When the cell confluence reached 80%-90%, the cells were collected, counted, and seeded. PBS containing 5×10.sup.6 cells was mixed with 100 uL Matrigel (final volume of 200 μL). The mouse was injected with the mixture on the right back side, 5×10.sup.6 cells/mouse. The mice were grouped when the tumor grew to a volume of 150-200 mm.sup.3 and administered intraperitoneally with the test sample three times per week. The tumor diameter was measured with a vernier caliper three times per week and the tumor volume was calculated by a calculation equation of 0.5a×b.sup.2, where a and b represent the long and short diameters of the tumor, respectively. The anti-tumor efficacy of antibodies was evaluated by the relative tumor growth rate T/C (%). The calculation equation of the relative tumor growth rate T/C (%) is as follows: T/C %=TRTV/CRTV×100% (TRTV: treatment group RTV; CRTV: negative control group RTV). RTV=V21/VO, where VO is the tumor volume measured at the time of being grouped and administered (Day 0), and V21 is the tumor volume measured at the 21th day of administration (Day 21). The tumor volumes on the last day (Day 21) of the administration group and the vehicle group were analyzed by t-test using GraphPad Prism. The results are shown in Table 9.

TABLE-US-00009 TABLE 9 Anti-tumor test results in human gastric cancer NUGC-4 xenograft model p (vs. Isotype Antibody Dose (mg/kg) T/C (%) IgG4) Isotype IgG4 6 86.61 — AB06.12-4P 6 18.82 ** Hu26T-31-PE 3 12.24 *** Hu26T-31-PE 6 2.96 *** Hu34-39-PE 3 15.29 *** Hu34-39-PE 6 3.45 *** Comparing with Isotype IgG4, *** indicates p < 0.001; ** indicates p < 0.005.

[0309] The experimental results show that the antibodies of the present disclosure have a significant anti-tumor effect in NOD-SCID mice xenograft model inoculated with human gastric cancer NUGC-4 cells. Hu26T-31-PE and Hu34-39-PE at a dose of 3 mg/kg show equivalent tumor-suppressive effect as the reference antibody AB06.12-4P at a dose of 6 mg/kg, while Hu26T-31-PE and Hu34-39-PE at a dose of 6 mg/kg show a better tumor-suppressive effect than the reference antibody AB06.12-4P at a dose of 6 mg/kg. One week after drug withdrawal, tumor recurrence occurred in the group of reference antibody at a dose of 6 mg/kg, while no recurrence occurred in the group of Hu26T-31-PE or Hu34-39-PE (FIG. 6 and FIG. 7). It indicates that the anti-CD47 antibodies of the present disclosure have unexpectedly better effect on inhibiting tumor growth.

Example 12 Production of Anti-Human PD-L1 Antibodies

1. Animal Immunization

[0310] Experimental BALB/c and SJL mice and SD rats were immunized with mFc-tagged PD-L1 antigen protein (purchased from ACROBiosystems, Beijing) together with adjuvants.

[0311] The immune adjuvant was Freund's Adjuvant, Complete (SIGMA, F5881-10ML) for the first time, and Freund's Adjuvant, Incomplete (SIGMA, F5506-10ML) for the later stages. Different tagged PD-L1 protein samples were added dropwise to the adjuvant solution with vortex to mix thoroughly. The dosages of the adjuvant were referred to the instructions. After the mixture was mixed well to form a water-in-oil emulsion, mice or rats were immunized. The immunization protocol is shown in Table 10.

TABLE-US-00010 TABLE 10 Immunization protocol Group Strain Antigen Adjuvant Dose Route* 1 BALB/c PBS None 2 BALB/c PD-L1-mFc Freund's Adjuvant, Complete (First time) 50 μg i.m. Freund's Adjuvant, Inomplete 25 μg 3 SJL PBS None 4 SJL PD-L1-mFc Freund's Adjuvant, Complete (First time) 50 μg i.m. Freund's Adjuvant, Inomplete 25 μg 5 SD PBS None 6 SD PD-L1-mFc Freund's Adjuvant, Complete (First time) 100 μg  i.m. Freund's Adjuvant, Inomplete 50 μg * i.m.: intramuscular injection

2. Cell Fusion

[0312] The mouse spleen was aseptically removed and prepared into a cell suspension, and the cells were fused at the ratio of spleen cells: Sp2/0 cells=1:1. The fused cell suspension was transferred to 15 mL RPMI 1640 complete medium containing 20% FBS and then left at room temperature for 20 min. The fused cells were resuspended with RPMI 1640 medium containing 1×HAT, 1×BIOMYC3, and 20% FBS. The cell suspension was added to several 96-well cell culture plates at 100 μl/well to ensure that the cell volume per well was about 4×10.sup.4 cells/well, and the plates was placed in a 37° C. cell incubator. After 5 days, additional 100 μL of RPMI 1640 complete medium containing 20% FBS, 1×HAT, and 1×BIOMYC-3 was added to each well.

3. Screening of Positive Clones

[0313] After one week of fusion, the cell supernatant was collected. The hybridoma parent clones with binding and blocking abilities were screened by ELISA, expanded, tested for binding and blocking activities, and screened again for hybridoma positive cell lines with binding and blocking abilities. The positive cell lines were subcloned by the limiting dilution method, and after one week of culture, the activities of binding to PD-L1 molecules and blocking the interaction of PD-L1/PD-1 of the supernatant were detected by ELISA. Four preferable cell strains that showed positive results in the above two tests, PL-7, The PL-15, PL-16 and PL-18, were selected.

4. Acquisition of Variable Region Sequences of Anti-PD-L1 Antibodies

[0314] The subcloned positive hybridoma cells were expanded, and an appropriate amount of cells was used for total RNA extraction according to the instructions of RNeasy Plus Mini Kit (Qiagen, 74134). The first strand of cDNA was synthesized using Prime Script 1st strand cDNA Synthesis Kit (Takara, 6110A).

[0315] Specific primers were designed according to the variable region of the mouse antibody subtype, and 5′ end of the primers contained the homologous arm sequence for homologous recombination with the eukaryotic expression vector. PCR amplification for the variable region of antibodies was performed using cDNA as a template to obtain the gene fragments of the light chain variable region and heavy chain variable region of the mouse antibody, named SHS009PL-7, SHS009PL-15, SHS009PL-16 and SHS009PL-18, respectively. The design of primers refers to references: Anke Krebber, Susanne Bornhauser, Jorg Burmester et al. Reliable cloning of functional antibody variable domains from hybridomas and spleen cell repertoires employing a reengineered phage display system. Journal of Immunological Methods, 1997, 201: 35-55; 2. Simon KorenMihaKosmaĉAnjaColjaVenturinietal. Antibody variable-region sequencing as a method for hybridoma cell-line authentication, 2008, 78: 1071-1078. DNA sequencing was performed and the results are shown in Table 11.

TABLE-US-00011 TABLE 11 Sequence table of anti-PD-L1 murine-derived monoclonal antibody H2CDR1/ L2CDR1/ Amino acid Amino acid H2CDR2/ L2CDR2/ Antibody sequence of VH sequence of VL H2CDR3 L2CDR3 SHS009PL-7 SEQ ID NO: 99 SEQ ID NO: 103 SEQ ID NOs: SEQ ID NOs: 75, 76 and 77 78, 79 and 80 SHS009PL-15 SEQ ID NO: 100 SEQ ID NO: 104 SEQ ID NOs: SEQ ID NOs: 81, 82 and 83 84, 85 and 86 SHS009PL-16 SEQ ID NO: 101 SEQ ID NO: 105 SEQ ID NOs: SEQ ID NOs: 87, 88 and 89 90, 91 and 92 SHS009PL-18 SEQ ID NO: 102 SEQ ID NO: 106 SEQ ID NOs: SEQ ID NOs: 93, 94 and 95 96, 97 and 98

5. Construction of Chimeric Antibodies

[0316] The purified gene fragments of the light chain and the heavy chain variable regions of the mouse antibodies were respectively co-transformed into Escherichia coli DH5a competent cells with the linearized eukaryotic expression plasmid containing the light chain constant region or the heavy chain constant region of human antibodies. The chimeric antibodies with correct sequences were selected and named PL-7CHI, PL-15CHI, PL-16CHI and PL-18CHI. The sequencing results of the chimeric antibodies are shown in Table 12.

[0317] The amino acid sequences of VL and VH of PL-7CHI, PL-15CHI, PL-16CHI and PL-18CHI are identical to those of the murine-derived antibodies SHS009PL-7, SHS009PL-15, SHS009PL-16 and SHS009PL-18, respectively.

[0318] The plasmids for light and heavy chains of chimeric antibodies were transfected into HEK 293F cells, and the antibodies were expressed and purified. The purity, activity and affinity were tested and analyzed.

TABLE-US-00012 TABLE 12 Sequence table of anti-PD-L1 chimeric antibodies Amino Amino Chimeric antibody acid sequence of VH acid sequence of VL PL-7CHI SEQ ID NO: 99 SEQ ID NO: 103 PL-15CHI SEQ ID NO: 100 SEQ ID NO: 104 PL-16CHI SEQ ID NO: 101 SEQ ID NO: 105 PL-18CHI SEQ ID NO: 102 SEQ ID NO: 106

6. Humanization of Anti-PD-L1 Antibodies

[0319] Based on the results of activity analysis and affinity KD value of chimeric antibodies, PL-7CHI and PL-16CHI were modified to humanized antibodies.

[0320] The humanization of murine-derived monoclonal chimeric antibodies PL-7CHI and PL-16CHI was carried out with reference to the classical CDR transplantation strategy. With the framework region sequence FR1-FR3 of antibodies PL-7CHI and PL-16CHI as templates, full human framework regions with similar 3D structure but low immunogenicity were screened in the human framework region library to replace FR1-FR3 sequence of PL-7CHI and PL-16CHI. After homology comparison, it was found that the FR region sequences of the heavy chain variable region of the antibodies PL-7CHI and PL-16CHI were the most similar to the human antibody germline genes M99683IIGHV4-31*02 (SEQ ID NO: 107) and X62109IIGHV1-3*01 (SEQ ID NO: 108), respectively; the FR region sequences of the light chain variable region of the antibodies PL-7CHI and PL-16CHI were the most similar to the human antibody germline gene Z00023IIGKV4-1*01 (SEQ ID NO: 109). The full-length sequences of the humanized heavy/light chain were 3D modeled and compared structurally with the heavy/light chain sequences of the original antibodies. The antigenicity and 3D structural similarity were considered comprehensively, and the amino acids that were shown to play a key role in the structural stability of the antibody in the structural simulation were back mutated to murine amino acid residues. Finally, 5 humanized heavy chains (the humanized heavy chain variable region sequences of PL-7CHI: VH1-0 (SEQ ID NO: 110), VH1-1 (SEQ ID NO: 111), VH1-2 (SEQ ID NO: 112), VH1-3 (SEQ ID NO: 113) or VH1-4 (SEQ ID NO: 114)) and 4 humanized light chains (the humanized light chain variable region sequences of PL-7CHI: VL1-0 (SEQ ID NO:115), VL1-1 (SEQ ID NO:116), VL1-2 (SEQ ID NO:117) or VL1-3 (SEQ ID NO:118)) of PL-7CHI were obtained; and 5 humanized heavy chains (the humanized heavy chain variable region sequences of PL-16CHI: VH1-0 (SEQ ID NO: 119), VH1-1 (SEQ ID NO: 120), VH1-2 (SEQ ID NO: 121), VH1-3 (SEQ ID NO: 122) or VH1-4 (SEQ ID NO: 123)) and 3 humanized light chains (the humanized light chain variable region sequences of PL-16CHI: VL1-0 (SEQ ID NO: 124), VL1-1 (SEQ ID NO: 125) or VL1-2 (SEQ ID NO: 126)) of PL-16CHI were obtained. On this basis, multiple humanized antibodies were obtained through different combinations of light and heavy chains. After activity detection, it was determined that the anti-PD-L1 humanized antibodies with the highest scores were HuPL7-21 and HuPL16-42.

Example 13 Production of Humanized Anti-PD-L1 Antibodies

[0321] Corresponding polynucleotides were synthesized based on the amino acid sequences of the light chain and heavy chain variable region of humanized antibody obtained above, and oligonucleotide fragments containing complementary sequences between adjacent fragments were synthesized. The oligonucleotide fragments were annealed and assembled by Overlap PCR. Then nucleotide fragments encoding the entire light chain and heavy chain variable regions were amplified by specific primers (5′ end contained the homologous arm sequence for homologous recombination with the eukaryotic expression vector). The purified nucleotide fragments of the light chain variable region were co-transformed into Escherichia coli DH5a competent cells with the linearized eukaryotic expression plasmid containing light chain constant region of IgG4. The purified nucleotide fragments of the heavy chain variable region were co-transformed into Escherichia coli DH5a competent cells with the eukaryotic expression plasmid containing heavy chain constant region of IgG4 containing S228P/L235E mutation. The competent cells with the transformed plasmids were spread evenly on the surface of the agar plates containing the corresponding antibiotics. The agar plates were incubated in a 37° C. constant temperature incubator overnight, and then several single colonies were picked out for DNA sequencing.

[0322] The positive clones with correct sequences were subjected to plasmid extraction to obtain the expression plasmids for the light chain and the heavy chain of humanized antibodies. The concentration and purity of the plasmids were detected by a nucleic acid quantitative analyzer.

[0323] The plasmids were transfected into HEK293 F cells, and a large number of antibodies were expressed and purified. The purity, activity and affinity were tested and analyzed. The sequences are shown in Table 13.

TABLE-US-00013 TABLE 13 Sequence table of anti-PD-L1 humanized antibodies Humanized Amino acid sequence Nucleotide sequence antibody VH VL VH VL HuPL7-21 SEQ ID SEQ ID NO: 116 SEQ ID SEQ ID NO: NO: 112 NO: 127 129 HuPL16-42 SEQ ID SEQ ID NO: 126 SEQ ID SEQ ID NO: NO: 123 NO: 128 130

[0324] For the antibody HuPL7-21, the sequences of CDR1, CDR2 and CDR3 of VH are set forth in SEQ ID NOs: 75, 76 and 77, respectively; the sequences of CDR1, CDR2 and CDR3 of VL are set forth in SEQ ID NOs: 78 and 79 and 80, respectively. For the antibody HuPL16-42, the sequences of CDR1, CDR2 and CDR3 of VH are set forth in SEQ ID NOs: 87, 88 and 89, respectively; the sequences of CDR1, CDR2 and CDR3 of VL are set forth in SEQ ID NOs: 90, 91 and 92, respectively.

Example 14 Construction, Expression and Purification of Anti-CD47/PD-L1 Bispecific Antibodies

1. Construction of Expression Vector for Bispecific Anti-CD47/Anti-PD-L1 Antibodies

[0325] The anti-CD47/anti-PD-L1 bispecific antibodies were constructed by genetic engineering, and the structure of antibodies is shown in FIG. 8. The bispecific antibody is formed by heterodimerization of two single chains from anti-PD-L1 antibody and anti-CD47 antibody, respectively. Unlike the natural IgG antibody, the light chains of both anti-PD-L1 and anti-CD47 antibodies in this bispecific antibody are linked to the N-terminal of the heavy chain by an additional flexible linking peptide, which is a GGGGS repeat sequence containing glycine (G) and serine (S) residues, preferably a sequence containing eight GGGGS repeats. In addition, in order to promote the formation of heterodimers, the S354C/T366W mutation is further added to the CH3 domain of the anti-PD-L1 antibody single chain and the Y349C/T366S/L368A/Y407V mutation to the CH3 domain of the anti-CD47 antibody single chain on the basis of the above S228P/L235E mutation. According to this structural form, the sequences of the above-mentioned anti-PD-L1 antibody HuPL7-21 (light chain variable region set forth in SEQ ID NO: 116 and constant region CL set forth in SEQ ID NO: 131) or HuPL16-42 (light chain variable region set forth in SEQ ID NO: 126 and constant region CL set forth in SEQ ID NO: 131) were used to obtain the nucleotide fragment of the anti-PD-L1 antibody single chain by gene synthesis, namely ScFabHuPL7-21Ks or ScFabHuPL16-42Ks. The sequences of the above-mentioned anti-CD47 antibody Hu34-39 (light chain variable region set forth in SEQ ID NO: 33 and constant region CL set forth in SEQ ID NO: 131) were used to obtain the nucleotide fragment of the anti-CD47 antibody single chain by gene synthesis, namely ScFabHu34-39Hs. The nucleotide fragments ScFabHuPL7-21Ks (or ScFabHuPL16-42Ks) and ScFabHu34-39Hs (upstream and downstream containing homology arms of appropriate length) were respectively co-transformed into Escherichia coli DH5a competent cells with the linearized eukaryotic expression plasmid pHR. The competent cells transformed with the plasmids were spread evenly on the surface of the agar plates containing the corresponding antibiotics. The agar plates were incubated in a 37° C. constant temperature incubator overnight, and then several single colonies were picked out for DNA sequencing. The positive clones with correct sequence were subjected to plasmid extraction to obtain ScFabHuPL7-21Ks (or ScFabHuPL16-42Ks) and ScFabHu34-39Hs expression vectors.

[0326] FIG. 9 shows a schematic diagram of the sequence structure of ScFabHuPL7-21Ks, wherein HuPL7-21VL-CL represents humanized PD-L1 monoclonal antibody HuPL7-21 light chain; (GGGGS)8 represents a flexible linking peptide of 8 GGGGS repeats; HuPL7-21VH represents humanized PD-L1 monoclonal antibody HuPL7-21 heavy chain variable region; IgG4CH/Ks represents IgG4 heavy chain constant region (specific sequence set forth in, for example, SEQ ID NO: 132) containing S228P/L235E/S354C/T366W mutation to form a “Knob” structure.

[0327] FIG. 10 shows a schematic diagram of the sequence structure of ScFabHu34-39Hs, wherein Hu34-39VL-CL represents humanized CD47 monoclonal antibody Hu34-39 light chain; (GGGGS)6 represents a flexible linking peptide of 6 GGGGS repeats; Hu34-39VH represents humanized CD47 monoclonal antibody Hu34-39 heavy chain variable region; IgG4CH/Hs represents IgG4 heavy chain constant region (specific sequence set forth in, for example, SEQ ID NO: 133) containing S228P/L235E/Y349C/T366S/L368A/Y407V mutation to form a “Hole” structure.

[0328] The CL sequence in the above structure is set forth in SEQ ID NO: 131.

2. Transient Expression of Anti-CD47/Anti-PD-L1 Bispecific Antibody in Expi-CHO Cells

[0329] Expi-CHO cells were transfected with the recombinant plasmids of the above two expression vectors ScFabHuPL7-21Ks (or ScFabHuPL16-42Ks) and ScFabHu34-39Hs using the Expi-Fectamine CHO Transfection Kit. After culturing in serum-free medium for 14 days, the supernatant of Expi-CHO cells was collected and detected the expression of the bispecific antibody by Western blotting. The bispecific antibody formed by dimerization of two single chains of ScFabHuPL7-21Ks and ScFabHu34-39Hs was named ScFab (HuPL7-21Ks/Hu34-39Hs), and the bispecific antibody formed by dimerization of two single chains of ScFabHuPL16-42Ks and ScFabHu34-39Hs was named ScFab (HuPL16-42Ks/Hu34-39Hs).

3. Purification of Bispecific Anti-CD47/Anti-PD-L1 Antibodies

[0330] After expressed and secreted in Expi-CHO cells, the bispecific antibodies of the present disclosure were purified by the method of Protein A affinity chromatography with the following specific steps. After the Protein A affinity chromatography column was equilibrated with buffer, the supernatant of Expi-CHO cell culture concentrated by ultrafilter was injected, monitored at A280 (nm), washed with washing solution until the unbound protein was all washed away, and then the antibodies were eluted with elution buffer to obtain the corresponding bispecific antibodies. The purified bispecific antibodies were tested for purity by SEC-HPLC and for molecular weight by LC-MS, and subjected to quality identification for subsequent pharmaceutical research. The identification results of SEC-HPLC and LC-MS showed that the purities of the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) both reached more than 95%, and the determination value of molecular weight matched the theoretical value.

Example 15 Construction of Stable Cell Line with High Expression of hPD-L1

[0331] Construction of Stable Cell Lines CHO-K1-hPD-L1 and Raji-hPD-L1 with High Expression of hPD-L1

[0332] CHO-K1 cells (from Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences) and Raji cells (from ImmuneOnco Biopharmaceuticals Inc.) were transfected with the eukaryotic expression plasmid pTargeT-hPD-L1 containing hPD-L1 (human PD-L1) extracellular region sequence (UniProtKB-Q9NZQ7(PD-L1 Human)>splQ9NZQ7119-238, SEQ ID NO: 134) by electroporation, and cultured in an incubator at 37° C. and 5% CO.sub.2. After 24 h, cells were selected with a medium containing 500 μg/ml G418. After 12 days, the positive rate of the cell pool was detected by FACS. The cells after electroporation were plated (at a cell density of 1×10.sup.6 cells/ml, 100 μl/well) and incubated with FITC anti-human PD-L1 antibody (SINO BIOLOGICAL, 10084-MMB6-F) at 4° C. for 60 min. The mean value under the FITC channel was read by a flow cytometer. After data analysis, the positive cell lines were selected for subcloning, and the CHO-K1/Raji cells from a single clone were selected. The cell lines with high expression of PD-L1 were named CHO-K1-hPD-L1 and Raji-hPD-L1.

Example 16 In Vitro Binding and Blocking Tests of Anti-CD47/Anti-PD-L1 Bispecific Antibody (by ELISA)

1. Anti-CD47/Anti-PD-L1 Bispecific Antibody Binding to PD-L1 (by ELISA)

[0333] Human PD-L1-His protein (0.5 μg/ml, 100 μl/well) was coated on a 96-well ELISA plate and incubated at 37° C. for 2 h. After washing 3 times with 1×PBST, the plate was blocked with 5% nonfat milk at 4° C. overnight and then washed 3 times with 1×PBST. The bispecific antibody was added to the ELISA plate with concentrations starting from 10 μg/mL at 5-fold serial dilution, and incubated at 37° C. for 1.5 h, with Atezolizumab (Sino Biological, Cat: 68049-H001, abbreviated Ate) as a control antibody. After washing 5 times with 1×PBST, HRP-Anti-Human IgG secondary antibody (Jackson, 109-035-003, 1:10000) was added and incubated at 37° C. for 40 min. After washing 5 times with 1×PBST, color developing solution TMB was added to the plate, and a microplate reader (Thermo, Multiskan FC) was used to read OD450 value after termination of the reaction. The EC.sub.50 results are shown in FIG. 11.

[0334] The experimental results show that both the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) are able to bind to human PD-L1, and the binding to human PD-L1 was comparable to that of Atezolizumab.

2. Anti-CD47/Anti-PD-L1 Bispecific Antibody Competing with PD-1 (by ELISA)

[0335] The activity of antibodies to block the binding of PD1 to PD-L1 was analyzed by ELISA. Human PD-L1-hFC protein (2 μg/ml, 100 μl/well) was coated on a 96-well ELISA plate and incubated at 37° C. for 2 h. After washing 3 times with 1×PBST, the plate was blocked with 5% nonfat milk at 4° C. overnight and then washed 3 times with 1×PBST. The bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) or ScFab (HuPL16-42Ks/Hu34-39Hs) as the primary antibody was added to the ELISA plate with a total of 8 concentrations starting from 10 μg/mL at 3-fold serial dilution, and incubated together with 1 μg/mL PD-1-mFc at 37° C. for 1.5 h in the presence of serially diluted anti-PD-L1 antibody, with Atezolizumab (Sino Biological, Cat: 68049-H001, abbreviated Ate) as a control antibody. After washing 5 times with 1×PBST, Anti-Mouse IgG HRP (Jackson, 109-035-003, 1:10000) was used as the secondary antibody and incubated at 37° C. for 1 h. After washing 5 times with 1×PBST, color developing solution TMB was added to the plate, and the microplate reader was used to read OD450 value after termination of the reaction. The results are shown in FIG. 11.

[0336] The experimental results show that both the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) had the ability to block the binding of PD-L1 to PD-1, and the blocking ability was comparable to that of Atezolizumab.

3. Anti-CD47/Anti-PD-L1 Bispecific Antibody Binding to CD47 (by ELISA)

[0337] The binding activity of antibodies to CD47 was analyzed by ELISA. Human CD47-His protein (0.5 μg/ml, 100 μl/well) was coated on a 96-well ELISA plate and incubated at 37° C. for 2 h. After washing 3 times with 1×PBST, the plate was blocked with 5% nonfat milk at 4° C. overnight and then washed 3 times with 1×PBST. The bispecific anti-CD47/anti-PD-L1 antibody provided by the present disclosure was added to the ELISA plate with a total of 8 concentrations starting from 50 μg/mL at 5-fold serial dilution, and incubated at 37° C. for 1.5 h, with Hu34-39-PE as a control antibody. After washing 5 times with 1×PBST, HRP-Anti-Human IgG (1:10000) was used as the secondary antibody and incubated at 37° C. for 40 min. After washing 5 times with 1×PBST, color developing solution TMB was added to the plate, and a microplate reader (Thermo, Multiskan FC) was used to read OD450 value after termination of the reaction. The EC.sub.50 results are shown in FIG. 11.

[0338] The experimental results show that both the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) are able to bind to human CD47, and their binding activities were reduced to varying degrees. The binding EC.sub.50 value of ScFab (HuPL7-21Ks/Hu34-39Hs) was 40-fold higher than that of Hu34-39-PE.

4. Anti-CD47/Anti-PD-L1 Bispecific Antibody Competing with SIRPα (by ELISA)

[0339] The blocking activity of antibodies was analyzed by ELISA. Human CD47-His protein (0.4 μg/ml, 100 μl/well) was coated on a 96-well ELISA plate and incubated at 37° C. for 2 h. After washing 3 times with 1×PBST, the plate was blocked with 5% nonfat milk at 4° C. overnight and then washed 3 times with 1×PBST. The bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) or ScFab (HuPL16-42Ks/Hu34-39Hs) as the primary antibody was added to the ELISA plate with a total of 8 concentrations starting from 10 μg/mL at 3-fold serial dilution, and incubated together with 2 μg/mL SIRPα-biotin at 37° C. for 1.5 h in the presence of serially diluted anti-CD47 antibody, with Hu34-39-PE as a control antibody. After washing 5 times with 1×PBST, SA-HRP (Jackson, 109-035-003, 1:10000) was used as the secondary antibody and incubated at 37° C. for 1 h. After washing 5 times with 1×PBST, color developing solution TMB was added to the plate, and the microplate reader was used to read OD450 value after termination of the reaction. The results are shown in FIG. 11.

[0340] The experimental results show that both the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) are able to block the binding of CD47 to SIRPα, and the blocking ability was lower than that of Hu34-39-PE, with an IC.sub.50 value increased by about 40 times.

5. Anti-CD47/Anti-PD-L1 Bispecific Antibody Competing with CD80 (by ELISA)

[0341] The activity of antibodies to block the binding of PD-L1 to CD80 was analyzed by ELISA. CD80-hFc protein (8 μg/ml, 100 μl/well) was coated on a 96-well ELISA plate and incubated overnight. After washing 3 times with 1×PBST, the plate was blocked with 5% nonfat milk at 37° C. for 2 h and then washed 3 times with 1×PBST. The bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) or ScFab (HuPL16-42Ks/Hu34-39Hs) as the primary antibody was added to the ELISA plate with a total of 8 concentrations starting from 30 μg/mL at 3-fold serial dilution, and incubated together with PD-L1-mFc at 37° C. for 1.5 h in the presence of serially diluted anti-PD-L1 antibody, with Atezolizumab as a control antibody. After washing 5 times with 1×PBST, Anti-Mouse IgG HRP was used as the secondary antibody and incubated at 37° C. for 1 h. After washing 5 times with 1×PBST, color developing solution TMB was added to the plate, and the microplate reader was used to read OD450 value after termination of the reaction. The results are shown in FIG. 11

[0342] The experimental results show that both the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) are able to block the binding of PD-L1 to CD80, and the blocking ability was comparable to that of Atezolizumab.

[0343] The above experimental results show that the bispecific antibodies of the present disclosure bind to PD-L1/CD47 with different activities, so as to reduce the toxic response of the antibodies such as blood toxicity while ensuring the anti-tumor activity.

Example 17 Binding/Blocking Test of Anti-CD47/Anti-PD-L1 Bispecific Antibody at Cellular Level

1. Binding Activity of Bispecific Antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) to PD-L1/CD47 on Cell Surface Detected by FACS

[0344] The CHO-K1-hPD-L1/CHO-K1-hCD47 stable cell line was used as a PD-L1/CD47 provider, and the serially diluted bispecific anti-PD-L1/anti-CD47 antibody was added to the cells as the primary antibody and incubated at 4° C. for 1.5 h. PE Anti-Human IgG was used as the secondary antibody and incubated at 4° C. for 1 h. Atezolizumab and Hu34-39-PE were used as positive controls. The product of the mean value and the Parent value at a wavelength of 585 nm was read using a flow cytometer, and the results are shown in FIG. 12.

[0345] The experimental results show that both the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) are able to bind to human PD-L1 on cell surface, and the binding ability was comparable to that of Atezolizumab. The binding activity of bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) to CD47 on cell surface was lower than that of Hu34-39-PE monoclonal antibody, with EC.sub.50 increased by about 4 times and Emax decreased by about 2 times.

2. Ability of Bispecific Antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) to Block the Binding of PD-1/SIRPα to Cell Surface Detected by FACS

[0346] The CHO-K1-hPD-L1/CHO-K1-hCD47 stable cell line was used as a PD-L1/CD47 provider to observe the binding of PD-L1 to PD-1 and the binding of CD47 to SIRPα in the presence of serially diluted anti-PD-L1 antibody/anti-CD47 antibody. The bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) or ScFab (HuPL16-42Ks/Hu34-39Hs) was used as the primary antibody, added to the cells after serial dilution together with 1 μg/mL PD-1-mFc and SIRPα-biotin respectively, and incubated at 37° C. for 1.5 h. PE-Anti-Mouse IgG/PE-SA was used as the secondary antibody. Atezolizumab was used as a positive control for blocking the binding of PD-1-mFc to PD-L1 on cell surface, and Hu34-39-PE was used as a positive control for blocking the binding of SIRPα to CD47 on cell surface. The results are shown in FIG. 12.

[0347] The experimental results show that both the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) are able to block the binding of human PD-1 to CHO-K1-PD-L1 and SIRPα to CHO-K1-CD47, and the ability to block the binding of PD-1 to CHO-K1-PD-L1 was comparable to that of Atezolizumab. The ability of the bispecific antibodies to block the binding of SIRPα to CHO-K1-CD47 was lower than that of Hu34-39-PE, with an IC.sub.50 value increased by 3 times.

[0348] According to the results of ELISA and FACS, the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) had the similar ability of binding to and blocking of PD-L1 as Atezolizumab. While the ability of binding to CD47 decreased, the EC.sub.50 increased by about 40 times detected by ELISA and the EC.sub.50 increased by about 4 times and Emax decreased by 2 times detected by FACS. That is, the bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) bind to PD-L1/CD47 with different activities, which help to enhance the tumor targeting and reduce the adverse reactions, especially those against red blood cells, of the bispecific antibodies.

Example 18 Double Binding and Double Blocking Tests of Anti-CD47/Anti-PD-L1 Bispecific Antibody on Raji-hPD-L1 Cells Detected by FACS

[0349] Raji-hPD-L1 tumor cells were purchased from ImmuneOnco Biopharmaceuticals Inc. Both of hPD-L1 and hCD47 were highly expressed on the surface of Raji-hPD-L1 cells. The bispecific antibodies ScFab (HuPL7-21Ks/Hu34-39Hs) and ScFab (HuPL16-42Ks/Hu34-39Hs) bound to both hPD-L1 and hCD47 on the surface of Raji-hPD-L1 cells, and also blocked the binding of CD47/SIRPα and the binding of PD-1/PD-L1 at the same time, showing dual-arm binding and dual-arm blocking activities. Dual-blocking: 2.4×10.sup.5 per well Raji-hPD-L1 cells were plated in a plate with U-bottom. For the addition of the primary antibody, the antibody was diluted in 2-fold serial dilution to a total of 9 working concentrations starting from 2.5 μg/mL; SIRPα-mFc with a final antigen concentration of 1 μg/mL was evenly mixed with PD-1-mFc with a final concentration of 1 μg/mL; and 50 μl of the antibody and 50 μl of the antigen were premixed in each well, added to cell wells with a total of 100 μl, and then incubated at 4° C. for 1.5 h. For the addition of the secondary antibody, 200 μl of cell stain buffer was added first to wash three times, and 0.8 μl of PE-anti-mouse-IgG-Fc was added to each well and incubated at 4° C. for 1 h. Finally, 200 μl of cell stain buffer was added to wash three times, and 100 μl of cell stain buffer was added to resuspend the cells. A flow cytometry was used for detection. Dual-binding: 1.5×10.sup.6 per well Raji-hPD-L1 cells were plated in a plate with U-bottom. For the addition of the primary antibody, the antibody was diluted in 5-fold serial dilution to a total of 8 working concentrations starting from 10 μg/mL, and then incubated at 4° C. for 1.5 h. For the addition of the secondary antibody, 200 μl of cell stain buffer was added first to wash three times, and 0.8 μl of PE-anti-human-IgG-Fc was added to each well and incubated at 4° C. for 1 h. Finally, 200 μl of cell stain buffer was added to wash three times, and 100 μl of cell stain buffer was added to resuspend the cells. A flow cytometry was used for detection.

[0350] The experimental results are shown in FIG. 13. The results show that the bispecific antibody of the present disclosure bound to cells with dual-target, and simultaneously blocked the binding of CD47/SIRPα and the binding of PD-1/PD-L1 on the surface of the cells.

Example 19 Detection of the Inhibition of Anti-CD47/Anti-PD-L1 Bispecific Antibody on the Growth of Transplanted Tumor in Mice

[0351] With the characteristics of NOD, Prkdcscid, IL2rgnull deletion/mutation, NSG mice are the tool mice with the highest degree of immunodeficiency and are most suitable for human cell transplantation, showing little rejection of human cells and tissues. In the present disclosure, NSG mice (purchased from Biocytogen Pharmaceuticals (Beijing) Co., Ltd.) and Raji-hPD-L1 tumor cells were used to establish a tumor xenograft model, Raji-PBMC-NSG model, to study the antitumor effect of the anti-CD47/anti-PD-L1 bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) in a subcutaneous transplant model of Raji-hPD-L1 lymphoma. The anti-CD47 monoclonal antibody 5F9 was purchased from Sino Biological Inc. (Cat: 68063-H001). Table 14 shows the experimental design of the antitumor effect of the test drugs in Raji-PBMC-NSG tumor model.

TABLE-US-00014 TABLE 14 Experimental design for drug test in Raji-PBMC-NSG model Dosing Mouse Dosage volume Route of Dosing Group N Antibody strain (mg/kg) (μl/g) administration frequency 1 6 PBS NSG — 10 i.v. q.w × 3 2 6 Atezolizumab NSG 10 10 i.v. q.w × 3 3 6 Atezolizumab + 5F9 NSG 10 + 10 10 i.v. q.w × 3 4 6 BiAb NSG 10 10 i.v. q.w × 3 5 6 BiAb NSG 20 10 i.v. q.w × 3 a: N refers to the number of mice per group; b: Bi Ab refers to bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs).

[0352] The changes of tumor volume in each group over time are shown in FIG. 14. The anti-tumor effect of the bispecific antibody of the present disclosure was significantly superior to that of Atezolizumab, and was better than the combination of monoclonal antibodies (Atezolizumab+5F9).

Example 20 Acute Toxicity Test of Anti-CD47/Anti-PD-L1 Bispecific Antibody in Cynomolgus Monkey

[0353] In this example, a toxicity test of the anti-CD47/anti-PD-L1 bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) of the present disclosure (abbreviated as BiAb) was performed by administering to cynomolgus monkeys by single intravenous infusion. The administration doses of the bispecific antibody were 10, 30 and 100 mg/kg, one male and one female cynomolgus monkeys in each group. The anti-CD47 monoclonal antibody Hu34-39-PE was administered at a dose of 30 mg/kg, and Hu5F9 was administered at a dose of 20 mg/kg, with two cynomolgus monkeys in each group.

[0354] Monkeys were administered by single intravenous infusion, with an observation period of 21 days. Blood samples were collected from the femoral vein at different time points for the detection of blood cell count, coagulation function indexes, and blood biochemical indexes.

[0355] The results of drug safety evaluation show that as of the 21st day, no monkeys died in all groups, and there were no abnormalities in general state observation, food intake, body weight, etc. in each group.

[0356] Animals given the bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) showed no change in RBC count, HGB content and RET %. The bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) did not show toxicity on red blood cells and other hematological toxicity at doses of 10, 30 and 100 mg/kg.

[0357] Animals given the monoclonal antibody Hu5F9 (20 mg/kg) showed a significant decrease in RBC count and HGB content, and a significant increase in RET %. Animals given the monoclonal antibody Hu34-39-PE (30 mg/kg) showed a certain extent of decrease in RBC count and HGB content, and an increase in RET % less than the monoclonal antibody Hu5F9. Hu5F9 showed obvious toxicity on red blood cells, and Hu34-39-PE showed a weaker toxicity on red blood cells than Hu5F9. The safety of the bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) for red blood cells was significantly better than that of monoclonal antibodies Hu34-39-PE and Hu5F9.

[0358] The above examples show that the anti-CD47/anti-PD-L1 bispecific antibody ScFab (HuPL7-21Ks/Hu34-39Hs) of the present disclosure differentially binds to CD47 and PD-L1, and fairly retains the activities of binding to and blocking of PD-L1, without toxicity on red blood cells and other hematological toxicity, exhibiting an excellent safety.

Example 21 Detection of the Inhibition of Anti-PD-L1 Antibody on the Growth of Transplanted Tumor in Mice

[0359] In the present disclosure, NSG mice (purchased from Biocytogen Pharmaceuticals (Beijing) Co., Ltd., China) and Raji-hPD-L1 tumor cells (purchased from ImmuneOnco Biopharmaceuticals Inc., China) were used to establish a tumor xenograft model, Raji-PBMC-NSG model, to study the antitumor effect of the antibody of the present disclosure in a subcutaneous transplant model of Raji-hPD-L1 lymphoma. The anti-PD-L1 positive control antibody was Atezolizumab (Sino Biological, Cat: 68049-H001). There were 6 mice in each group. The negative control group was given normal saline (PBS). The anti-PD-L1 antibody HuPL7-21 of the present disclosure and Atezolizumab were administered at a dose of 10 mg/kg, respectively. The route of administration was intraperitoneal injection, and the dosing frequency was two times a week for three consecutive weeks.

[0360] At the end of the experiment, the tumor inhibition rate (TGI.sub.TV) % of each group was calculated, as shown in Table 15.

TABLE-US-00015 TABLE 15 Effect of each antibody on the tumor volume of Raji-PBMC-NSG model mice Group Dosage (mg/kg) TGI.sub.TV (%) PBS —  0.00 Atezolizumab 10 37.40 HuPL7-21 10 48.75

[0361] The anti-PD-L1 antibody HuPL7-21 of the present disclosure has a significantly better anti-tumor effect in vivo than Atezolizumab.

[0362] Although the present disclosure has been described above in detail, those skilled in the art should understand that various modifications and changes can be made to the present disclosure without departing from the spirit and scope of the present disclosure. The scope of the present invention should not be limited to the detailed description above, but should be attributable to the appended claims.