Anti-PD-L1 antibody coupling medicine and preparation method thereof

Abstract

The invention discloses an anti-PD-L1 antibody coupling drug and a preparation method thereof, belongs to the field of biological medicines, and is used for preparing anti-cancer drugs. The antibody coupling medicine is formed by coupling an anti-PD-L1 antibody and a payload through a linker, a light chain variable region sequence of the antibody is a sequence recorded by SEQ ID NO: 1, a heavy chain variable region sequence of the antibody is a sequence recorded by SEQ ID NO: 2, the payload is an avixatecan derivative Dxd, and the ratio (DAR) of the payload to the antibody is 6-8. The antibody-conjugated drug has a better killing effect on PD-L1 positive tumor cells of human and dogs, and has fewer dimers and higher thermal stability at a high DAR value (DAR6-8), and an accelerated stability test shows that the antibody-conjugated drug has good structural stability and binding activity.

Description

Technical Field

[0001] The present invention belongs to the field of biomedicine, and particularly relates to an anti-PD-L1 antibody conjugate drug and a preparation method thereof. Background Art

[0002] Antibody-drug conjugate (ADC) is a novel drug that combines the targeting of an antibody and the killing power of a small molecule drug (also called payload). The two are conjugated through a linker to form a "biological missile" structure, achieving precise delivery and high efficacy. In an antibody-drug conjugate, the antibody is responsible for specifically recognizing the disease target and guiding the drug to accurately locate at the disease site; the small molecule drug is usually a cytotoxin that can interfere with and inhibit the growth and proliferation of cancer cells or induce apoptosis.

[0003] The "Technical Guidelines for Non-clinical Research of Antibody-drug Conjugates" issued by the National Medical Products Administration (NMPA) stipulates that the in vitro stability of ADC needs to be investigated before the clinical trial is carried out. DAR (Drug-to-Antibody Ratio) is an important parameter of an antibody-drug conjugate, which refers to the average number of small molecule drugs conjugated to each antibody molecule. It is one of the core parameters in the design and development of antibody-drug conjugates, directly affecting the efficacy, toxicity and side effects, and stability of the drug. Generally, it is considered that when designing an antibody-drug conjugate, if the selected DAR is too low, the cytotoxicity is insufficient; if it is too high, the drug may be recognized by the immune system and is more likely to aggregate.

[0004] Cysteine is a commonly used conjugation site for antibody-drug conjugates. Currently, it is generally believed that when conjugating the drug payload through cysteine residues, the overall conformation of the antibody does not change significantly, but the energy of its unfolded conformation decreases, and the local surface hydrophobicity increases, and the melting temperature (Tm) of the antibody decreases accordingly, which may cause aggregation. Antibody-drug conjugates with a high DAR value are more likely to aggregate, and the presence of high DAR components in the mixture will determine the stability of the entire ADC. When DAR is 8, the hydrophobicity of its drug payload may directly cause the instability of monoclonal antibody (mAb). For cysteine-conjugated ADCs, with the increase of DAR or ionic strength, its thermal stability decreases significantly (Reference: Duerr C, Friess W. Antibody-drug conjugates - stability and formulation. Eur J Pharm Biopharm. 2019 Jun;139:168-176).

[0005] In the past, the research and development of antibody-drug conjugates has mainly focused on mature targets such as HER2, EGFR, Trop2, etc. However, immune checkpoints (such as PD-L1) have gradually become emerging research hotspots due to their unique biological characteristics. The development of antibody-drug conjugates requires systematic consideration of antibody characteristics. Different antibodies or engineering modifications may significantly affect their performance - glycosylation modification can change hydrophilicity, while changes in the flexible sequence of the hinge region or Fc mutations affect stability maintenance and change Fc functional effects at the same time. Antibodies with different CDR regions, conjugation methods, and DAR values may all affect key quality attributes such as the activity and stability of antibody-drug conjugates. For the same target, it is necessary to design according to the pharmacological mechanism and comprehensively consider the properties of ADCs with different technical solutions, and determine the drug structure and preparation method based on the consideration results. Summary of the Invention

[0006] The present invention provides an anti-PD-L1 antibody-drug conjugate and its preparation method, which is intended to be used for preparing anti-cancer drugs and has better stability at high DAR values. The technical solution is as follows: An anti-PD-L1 antibody-drug conjugate, whose molecule is composed of an anti-PD-L1 antibody and a payload (usually a small molecule drug) conjugated through a linker. The variable region sequence of the antibody light chain is the sequence recorded in SEQ ID NO:1, and the variable region sequence of the heavy chain is the sequence recorded in SEQ ID NO:2. And the ratio of the payload to the antibody within a single molecule (commonly known as the drug-to-antibody ratio, DAR) is 6 to 8, preferably the DAR is 8.

[0007] For the above anti-PD-L1 antibody-drug conjugate, the Fc segment of the antibody is of IgG1 type.

[0008] For the above anti-PD-L1 antibody-drug conjugate, the Fc segment of the antibody has N-linked glycosylation modification.

[0009] For the above anti-PD-L1 antibody-drug conjugate, the light chain sequence of the antibody is the sequence recorded in SEQ ID NO:3, and the heavy chain sequence is the sequence recorded in SEQ ID NO:4.

[0010] For the above anti-PD-L1 antibody-drug conjugate, the payload is exatecan derivative Dxd (Exatecanderivative for ADC). Dxd is a DNA topoisomerase I inhibitor, with the molecular formula C 26 H 24 FN3O6, with a molecular weight of 493.48, and its structural formula is as Figure 1 shown.

[0011] The above-mentioned anti-PD-L1 antibody conjugate drug, wherein the payload is conjugated to the anti-PD-L1 antibody through a cysteine residue.

[0012] The above-mentioned anti-PD-L1 antibody conjugate drug, wherein the linker is maleimide-GGFG peptide.

[0013] A method for preparing the above-mentioned anti-PD-L1 antibody conjugate drug, characterized by comprising the following steps: First step: Mix the above-mentioned anti-PD-L1 antibody with 20-fold molar amount of TCEP (tris(2-carboxyethyl)phosphine), and reduce at 37 °C for 2 hours; Second step: Add the Dxd-Linker conjugate according to the antibody:drug molar ratio of 1:10 to 1:15, and conjugate at 37 °C for 4-5 hours; desalt and purify to obtain the anti-PD-L1 antibody conjugate drug.

[0014] In the above steps, the anti-PD-L1 antibody is dissolved in phosphate buffer at pH 7.2 at a concentration of 10 mg / ml, and Dxd is dissolved in anhydrous DMSO at a concentration of 10 mmol / L.

[0015] The above-mentioned anti-PD-L1 antibody conjugate drug is used for the treatment of malignant tumors such as lung cancer, liver cancer, melanoma, etc.

[0016] The above-mentioned Dxd-Linker conjugate is MC-GGFG-DXD (English name Deruxtecan), wherein the toxic drug is Dxd, the linker is maleimide-GGFG peptide, provided by MedChemExpress (MCE) company, CAS number 1599440-13-7, catalog number HY-13631E, molecular formula C 52 H 56 FN9O 13 , molecular weight 1034.05, the structural formula is shown in Figure 22 . The antibody conjugate drug (ADC) constructed by MC-GGFG-DXD has Figure 23 the structure shown.

[0017] Compared with the conjugate drugs of foreign-listed antibodies, the above-mentioned anti-PD-L1 antibody conjugate drug has a better EC50 value (half maximal effective concentration) for human PD-L1, and has a better killing effect on human PD-L1 positive cells, and is used for the treatment of malignant tumors such as human lung cancer, liver cancer, melanoma, etc.

[0018] The above-mentioned anti-PD-L1 antibody conjugate drug has an affinity for canine PD-L1, and compared with the conjugate drugs of foreign marketed antibodies, it has a better killing effect on canine PD-L1 positive cells when DAR = 8, and is used for the treatment of canine malignant tumors, such as lung cancer, liver cancer, melanoma, etc.

[0019] The anti-PD-L1 monoclonal antibody conjugate drug (ADC) described in the present invention has fewer dimers at high DAR values (DAR6-8), can still maintain a binding activity equivalent to that of the naked antibody, the degree of improvement in cytotoxicity of ADC compared with free Dxd is significantly higher than that of traditional ADC drugs, and DSC method shows that the anti-PD-L1 monoclonal antibody conjugate drug described in the present invention has a higher thermal stability (Tm value) at DAR6-8 than at DAR2-4. The accelerated stability test shows that it still maintains good structural stability and binding activity after being stored in a liquid state at 2-8°C for two months at DAR6-8. Brief Description of the Drawings

[0020] Figure 1 : Chemical structural formula of Dxd.

[0021] Figure 2 : Determination results of the binding ability of the candidate humanized anti-PD-L1 monoclonal antibody to PD-L1 on the cell membrane surface.

[0022] Figure 3 : Determination results of the binding ability of the candidate humanized anti-PD-L1 monoclonal antibody to soluble PD-L1.

[0023] Figure 4 : Determination results of the ability of the candidate humanized anti-PD-L1 monoclonal antibody to competitively bind to PD-L1 on the cell membrane surface with Tecentriq.

[0024] Figure 5 : Determination results of the ability of the candidate humanized anti-PD-L1 monoclonal antibody to competitively bind to soluble PD-L1 with Tecentriq.

[0025] Figure 6 : Determination results of the ability of the candidate humanized anti-PD-L1 monoclonal antibody to competitively bind to PD-L1 on the cell membrane surface with PD-1.

[0026] Figure 7 : Determination results of the ability of the candidate humanized anti-PD-L1 monoclonal antibody to competitively bind to soluble PD-L1 with PD-1.

[0027] Figure 8 : Detection results of the biological activity of the candidate humanized anti-PD-L1 monoclonal antibody by the reporter gene method.

[0028] Figure 9 : Analysis results of the ADCC effect function of the candidate humanized anti-PD-L1 monoclonal antibody.

[0029] Figure 10 : Results of the CDC effector function analysis of the candidate humanized anti-PD-L1 monoclonal antibody.

[0030] Figure 11 : Results of the polyacrylamide gel electrophoresis analysis of anti-PD-L1 ADCs with different DAR values.

[0031] Figure 12 : Results of the isoelectric point analysis of anti-PD-L1 ADCs with different DAR values.

[0032] Figure 13 : Results of the purity analysis (SEC-HPLC method) of anti-PD-L1 ADCs with different DAR values.

[0033] Figure 14 : Results of the heterogeneity analysis (HIC-HPLC method) of anti-PD-L1 ADCs with different DAR values.

[0034] Figure 15 : Results of the detection of the binding activity of anti-PD-L1 ADC to soluble PD-L1 by ELISA.

[0035] Figure 16 : Results of the detection of the PD-L1 expression on the surface of A431 cells and MC38 cells.

[0036] Figure 17 : Results of the cytotoxicity detection of anti-PD-L1 ADC against A431 cells by CCK-8 method.

[0037] Figure 18 : Comparison of the results of the cytotoxicity detection of anti-PD-L1 ADC against MC38 cells and A431 cells by CCK-8 method.

[0038] Figure 19 : Results of the preliminary evaluation of the stability of anti-PD-L1 ADC by SDS-PAGE method.

[0039] Figure 20 : Results of the preliminary evaluation of the stability of anti-PD-L1 ADC by SEC-HPLC method.

[0040] Figure 21 : Results of the preliminary evaluation of the stability of the binding activity of anti-PD-L1 ADC by ELISA.

[0041] Figure 22 : Chemical structure formula of the Dxd-Linker conjugate MC-GGFG-DXD.

[0042] Figure 23 : Schematic diagram of the structure of the antibody-Dxd conjugate.

[0043] Figure 24 : Results of detecting the binding activity of anti-PD-L1 ADC to soluble canine PD-L1 by ELISA method.

[0044] Figure 25 : Detecting the binding ability of anti-PD-L1 ADC to PD-L1 on the surface of canine cell membrane by flow cytometry.

[0045] Figure 26 : Results of detecting the cytotoxicity of anti-PD-L1 ADC to canine tumor cell PETCC2478 by CCK-8 method.

[0046] Figure 27 : Results of detecting the cytotoxicity of anti-PD-L1 ADC to canine tumor cell PETCC2477 by CCK-8 method. Specific implementation mode Example 1: Preparation of recombinant anti-PD-L1 monoclonal antibody T0004 1.1 Screening of murine anti-PD-L1 monoclonal antibody by mouse hybridoma method

[0047] Immunization: Immunize Bab / c mice intraperitoneally with recombinant human PD-L1-Fc fusion protein at 100 - 500 μg / ml for 3 times. 3 - 7 days before fusion, boost immunize by tail vein injection of recombinant human PD-L1-Fc fusion protein at 50 - 100 μg / animal / 200 μl.

[0048] Fusion: Fuse mouse myeloma cell line NS-1 (1 - 2x10 / ml) with splenocytes of immunized Bab / c mice, then place them in a 96-well cell culture plate and screen with complete medium containing HAT. Replace half of the medium after 3 - 5 days, and visible clones will form in about 2 weeks.

[0049] Identification of culture supernatant: Detect the culture supernatant in the 96-well plate by ELISA. Coat recombinant human PD-L1-Fc protein and irrelevant Fc fusion protein, add primary antibody: culture supernatant in the 96-well plate, and then add secondary antibody: anti-mouse-HRP. The single positive clone of PD-L1-Fc is murine anti-PD-L1 monoclonal antibody, and the double positive clone is murine anti-human Fc monoclonal antibody. Select single positive clones and culture them with complete medium containing HT.

[0050] Subcloning: Clone the hybridoma clones secreting murine anti-PD-L1 monoclonal antibody at 0.5 per well with complete medium, and select single clone wells for culture with complete medium.

[0051] Subtype detection: Detect the culture supernatant of the prepared murine anti-PD-L1 monoclonal antibody with a mouse subtype detection kit to determine the light and heavy chain subtypes.

[0052] The candidate clones obtained according to the above method were successively detected for their abilities to bind to PD-L1 on the cell membrane surface (flow cytometry), to bind to soluble PD-L1 (ELISA method), to bind to PD-L1 with affinity (surface plasmon resonance SPR method), to compete with PD-1 for binding to PD-L1 (ELISA method), and to compete with Tecentriq (generic name Atezolizumab) for binding to PD-L1 (ELISA method). Finally, clone numbered E66 was screened and sequenced.

[0053] Table 1 Detection results of the binding ability of E66 to PD-L1 on the cell membrane surface Clone Name EC50 (ng / ml) E66 274.7112

[0054] Table 2 Detection results of the binding ability of E66 to soluble PD-L1 (ELISA) Clone Name EC50 (ng / ml) Mouse anti PD-L1 mAb Monoclonal Antibody 1.471

[0055] Table 3 Detection results of the affinity of E66 to PD-L1 (Biacore) Clone Name ka (1 / Ms) Kd (1 / s) KD (M) Rmax (RU) Chi² (RU²) U-value Remarks E66 3.748E6 3.062E-6 8.170E-13 151.6 1.21 43 Beyond the Instrument Detection Range

[0056] Table 4 Detection results of the competition of E66 with PD-1 for binding to PD-L1 (ELISA) Clone Name EC50 (μg / ml) Whether Competitive E66 6.298 Competitive

[0057] Table 5 Detection results of the competition of E66 with Tecentriq for binding to PD-L1 (ELISA) Clone Name EC50 (μg / ml) Whether Competitive E66 4.917 Competitive Sequencing showed that the light chain variable region of the murine anti-PD-L1 monoclonal antibody E66 had the sequence of SEQ ID NO:5, and the heavy chain variable region had the sequence of SEQ ID NO:6.

[0058] 1.2 Humanization of the murine anti-PD-L1 monoclonal antibody E66

[0059] Select a human antibody sequence with a high homology to the light and heavy chains of the murine antibody as a template. According to the crystal structure of the murine antibody, determine the amino acids involved within 5 Å of the CDR region. Perform back mutations on the amino acids that are different between the murine antibody within 5 Å of the CDR region and the human template to determine the humanized antibody sequence. The humanized candidate sequences for the light chain variable region are the sequences of SEQ ID NO:1 and SEQ ID NO:7, and the humanized candidate sequence for the heavy chain variable region is the sequence of SEQ ID NO:2. The two candidate antibodies formed by combining the light and heavy chains are named T0004-BC-LH (light chain variable region SEQ ID NO:1, heavy chain variable region SEQ ID NO:2) and T0004-C-LH (light chain variable region SEQ ID NO:7, heavy chain variable region SEQ ID NO:2) respectively. Fuse the variable region with the constant region of the IgG1κ type antibody. Thus, the full-length sequence of the light chain of T0004-BC-LH is SEQ ID NO:3, and the full-length sequence of the heavy chain is SEQ ID NO:4. The full-length sequence of the light chain of T0004-C-LH is SEQ ID NO:8, and the full-length sequence of the heavy chain is SEQ ID NO:4. Synthesize the coding DNA sequence according to the amino acid sequence, insert it into the expression vector and transfect CHO cells to express and prepare a certain amount of T0004-BC-LH and T0004-C-LH antibody proteins, and perform comparative analysis of biological activities according to the following method.

[0060] 1.2.1 Comparison of the binding ability of candidate humanized anti-PD-L1 monoclonal antibodies to cell membrane surface PD-L1

[0061] Stably transfect the plasmid of the full-length human PD-L1 expression vector into CHO-K1 host cells, and obtain CHO-K1 engineered cells (CHO-K1 / hmPD-L1) that stably express PD-L1 on the cell membrane surface through pressure screening. Collect CHO-K1 / hmPD-L1 in the logarithmic growth phase, add the primary antibody: humanized anti-PD-L1 monoclonal antibody at 15 concentrations with a 2-fold ratio starting from 10000 ng / ml. The control is the commercial anti-PD-L1 monoclonal antibody Tecentriq (generic name Atezolizumab). Incubate on ice for 45 min to allow it to bind to PD-L1 on the surface of CHO-K1 / hmPD-L1 cells. Wash twice with 1% FBS PBS, add the secondary antibody: goat anti-human IgG(H+L) FITC fluorescently labeled secondary antibody, incubate on ice for 45 min, and wash twice with 1% FBS PBS. Resuspend the precipitate in 1% FBS PBS and analyze it with a flow cytometer. The abscissa (logarithmic) is the concentration of the sample, and the ordinate is the mean fluorescence intensity. The curve obtained from the measurement results is shown in Figure 2 , and calculate the half-maximal effective dose EC50 value as shown in the following table.

[0062] Table 6 Detection Results of the Binding Ability of the Candidate Humanized Anti-PD-L1 Monoclonal Antibody to Cell Membrane Surface PD-L1 Sample Name EC50 (ng / ml) T0004-BC-LH 336.20383 T0004-C-LH 324.08248 Tecentriq (Control) 318.10122

[0063] 1.2.2 Analysis of the Binding Ability of the Candidate Humanized Anti-PD-L1 Monoclonal Antibody to Soluble PD-L1

[0064] Principle: The recombinantly expressed PD-L1-Fc fusion protein is coated on the enzyme-linked immunosorbent assay (ELISA) plate. The humanized anti-PD-L1 monoclonal antibody can specifically bind to it, and the horseradish peroxidase (HRP)-labeled anti-human secondary antibody can bind to it. After adding the HRP substrate, a color reaction occurs, and the depth of the color is related to its affinity for PD-L1-Fc.

[0065] The experimental method is as follows: Coating: Dilute PD-L1-Fc with the coating solution to 10 μg / ml, add 100 μl / well to the ELISA plate, and coat at 37 °C for 1 - 2 hours.

[0066] Blocking: Discard the coating solution, add the blocking solution, 350 μl / well, and block at 37 °C for 1 - 2 hours or at 2 - 8 °C overnight.

[0067] Washing: Wash the plate 7 times with a plate washer.

[0068] Primary antibody: Dilute the humanized anti-PD-L1 monoclonal antibody and the control Tecentriq with PBS to a final concentration of 5000 ng / ml at 15 two-fold serial dilutions, add 100 μl / well, and react at 37 °C for 1 - 2 hours.

[0069] Washing: Discard the liquid in the ELISA plate and wash the plate 7 times with a plate washer.

[0070] Secondary antibody: Dilute goat anti-human kappa-HRP 1:8000 with the diluent according to the instruction manual, add 100 μl / well, and react at 37 °C for 1 hour.

[0071] Washing: Discard the liquid in the ELISA plate and wash the plate 7 times with a plate washer.

[0072] Color development: Add the TMB substrate solution to the ELISA plate, 50 μl / well, develop color in the dark for 1 - 10 minutes, add 50 μl / well of the stop solution, and mix well quickly.

[0073] Reading: Read the OD450nm with an ELISA reader, and use OD570nm as the reference wavelength.

[0074] Perform curve fitting with the sample concentration (abscissa, logarithm value) against the OD value (ordinate). The results are shown in Figure 3 , and the calculated EC50 values are shown in the following table.

[0075] Table 7 Detection Results of the Binding Ability of the Candidate Humanized Anti-PD-L1 Monoclonal Antibody to Soluble PD-L1 (ELISA) Sample Name EC50 (ng / ml) T0004-BC-LH 38.17 T0004-C-LH 54.65 Tecentriq (Control) 56.53

[0076] 1.2.3 Detection of the SPR Affinity Constant of the Candidate Humanized Anti-PD-L1 Monoclonal Antibody The recombinantly expressed PD-L1-Fc fusion protein was conjugated with biotin to form PD-L1-Fc -Biotin, which was then conjugated to the chip of the BIACORE analyzer. Using the candidate humanized anti-PD-L1 monoclonal antibody and the control Tecentriq as analytes, the kinetic parameters of macromolecular interactions were measured according to the instructions of the BIACORE analyzer. Finally, regeneration was performed using 10 mmol / L Gly or 10 mmol / L Gly + 3 mol / L MgCl2. The detection results are shown in the following table.

[0077] Table 8 Detection Results of the Affinity of the Recombinant Anti-PD-L1 Humanized Monoclonal Antibody to PD-L1 (Biacore) Sample Name Ka (1 / Ms) kd (1 / s) KD (M) Rmax (RU) Chi² (RU²) U-value T0004-BC-LH 4.415E+6 4.452E-5 1.008E-11 18.30 0.377 7 T0004-C-LH 3.881E+6 4.852E-5 1.250E-11 21.28 0.265 5 Tecentriq (Control) 5.183E+6 1.410E-4 2.720E-11 20.75 0.894 4

[0078] 1.2.4 Analysis of the Ability of the Candidate Humanized Anti-PD-L1 Monoclonal Antibody to Compete with Tecentriq for Binding to Membrane Surface PD-L1 Tecentriq is the world's first commercialized anti-PD-L1 humanized monoclonal antibody. The CHO-K1 engineering cells (CHO-K1 / hmPD-L1) constructed in 1.2.1 were cultured to the logarithmic growth phase and collected. The candidate humanized anti-PD-L1 monoclonal antibody at different concentrations was serially diluted 15-fold from 500 μg / ml, and added to the cell culture together with sub-saturating concentration of Tecentriq-FITC to compete for binding to cell surface PD-L1. After mixing, it was incubated on ice for 45 min and washed twice with 1% FBS PBS. The precipitate was resuspended in 1% FBS PBS and analyzed by flow cytometry. With the abscissa (logarithmic) being the concentration of the sample and the ordinate being the mean fluorescence intensity, the reaction curve is shown in Figure 4 。. The EC50 value was calculated and is shown in the following table.

[0079] Table 9 Detection Results of the Ability of the Candidate Humanized Anti-PD-L1 Monoclonal Antibody to Compete with Tecentriq for Binding to Membrane Surface PD-L1 Sample Name EC50 (μg / ml) Whether Competitive T0004-BC-LH 1.23916 Competitive T0004-C-LH 1.20865 Competitive Tecentriq (Control) 1.16779 Competitive

[0080] 1.2.5 Analysis of the Ability of the Candidate Humanized Anti-PD-L1 Monoclonal Antibody to Compete with Tecentriq for Binding to Soluble PD-L1 Principle: Biotin-labeled Tecentriq can specifically bind to the recombinant PD-L1-Fc fusion protein coated on the ELISA plate. Observe whether the candidate humanized anti-PD-L1 monoclonal antibody can inhibit the binding of PD-L1 to Tecentriq. Then add avidin labeled with horseradish peroxidase (Avidin-HRP). After adding the HRP substrate, a color reaction will occur. If there is competition, the curve will be an inverted "S" shape. If there is no competition, it will be a "straight line" that is basically the same as the reading of biotin-labeled Tecentriq. The specific experimental method is as follows: Coating: Dilute the recombinant PD-L1-Fc fusion protein to 10 μg / ml with the coating solution, add 100 μl / well to the ELISA plate, and coat at 37°C for 1 - 2 hours.

[0081] Blocking: Discard the coating solution, add the blocking solution, 350 μl / well, and block at 37°C for 1 - 2 hours or at 2 - 8°C overnight.

[0082] Washing: Wash the plate 7 times with a plate washer.

[0083] Primary antibody: Dilute the candidate humanized anti-PD-L1 monoclonal antibody to 390 μg / ml, mix it with 2×Tecentriq-Biotin in equal volume, and the concentration after mixing is 195 μg / ml. Then mix it with 1×Tecentriq-Biotin in equal volume and dilute it successively. Dilute it to 14 concentrations in a 2-fold dilution. Add the diluted product to the ELISA plate at 100 μl / well and react at 37°C for 1 hour.

[0084] Washing: Discard the supernatant and wash the plate 7 times with a plate washer.

[0085] Secondary antibody: Dilute Avidin-HRP 1:1000 according to the instructions, add 100 μl / well, and react at 37°C for 1 hour.

[0086] Washing: Discard the supernatant and wash the plate 7 times with a plate washer.

[0087] Color development: Add the TMB substrate solution to the ELISA plate, 50 μl / well, and develop color in the dark for 15 minutes. Add the stop solution to the ELISA plate, 50 μl / well, and mix quickly.

[0088] Reading: Read the OD450nm with an ELISA reader, and use 570nm as the reference wavelength.

[0089] Analysis of test results: Fit the curve of the sample concentration (abscissa, logarithmic value) against the OD value (ordinate). The results are shown in Figure 5 . Calculate the EC50 value as shown in the following table.

[0090] Table 10 Detection results of the competition binding of candidate humanized anti-PD-L1 monoclonal antibody and Tecentriq to soluble PD-L1 Sample Name EC50 (μg / ml) Whether Competitive T0004-BC-LH 1.489 Competitive T0004-C-LH 1.429 Competitive Tecentriq (Control) 1.494 Competitive

[0091] 1.2.6 Analysis of the ability of candidate humanized anti-PD-L1 monoclonal antibody to competitively bind to PD-L1 on the cell membrane surface with PD-1 Culture the CHO-K1 engineering cells (CHO-K1 / hmPD-L1) constructed in 1.2.1 until the logarithmic growth phase and collect them; Dilute the candidate humanized anti-PD-L1 monoclonal antibody from 10 μg / ml in a 2-fold dilution for 15 concentrations, and add it together with sub-saturating concentration of PD-1-Biotin to the cells in the logarithmic phase to competitively bind to PD-L1 on the cell membrane surface. After mixing, incubate on ice for 45 min. Wash twice with 1% FBS PBS, add Avidin-PE fluorescently labeled secondary antibody, incubate on ice for 45 min, and wash twice with 1% FBS PBS. Resuspend the precipitate in 1% FBS PBS and analyze it using a flow cytometer.

[0092] The abscissa (logarithmic) is the concentration of the sample, and the ordinate is the mean fluorescence intensity. Plot the reaction curve, as shown in Figure 6 . Calculate the EC50, as shown in the following table.

[0093] Table 11 Detection results of the ability of candidate humanized anti-PD-L1 monoclonal antibody to competitively bind to PD-L1 on the cell membrane surface with PD-1 Sample Name EC50 (ng / ml) Whether Competitive T0004-BC-LH 0.16444 Competitive T0004-C-LH 0.19668 Competitive Tecentriq (Control) 0.16056 Competitive

[0094] 1.2.7 Analysis of the ability of candidate humanized anti-PD-L1 monoclonal antibody to competitively bind to soluble PD-L1 with PD-1 Principle: Biotin-labeled recombinant PD-L1-Fc fusion protein can specifically bind to PD-1-Fc coated on the enzyme-linked immunosorbent assay (ELISA) plate. Observe whether the recombinant anti-PD-L1 humanized monoclonal antibody can inhibit the binding of PD-1 to its ligand PD-L1. Then add avidin labeled with horseradish peroxidase (Avidin-HRP). After adding the HRP substrate, a color reaction will occur. If there is competition, the curve will be an inverted "S" shape; if there is no competition, it will be a "straight line" that is basically the same as the reading of biotin-labeled PD-L1-Fc.

[0095] Experimental method: Coating: Dilute the recombinant PD-1-Fc fusion protein to 10 μg / ml with the coating solution, add 100 μl / well to the ELISA plate, and coat at 37 °C for 1 - 2 hours.

[0096] Blocking: Discard the coating solution, add the blocking solution, 350 μl / well, and block at 37 °C for 1 - 2 hours or block overnight at 2 - 8 °C.

[0097] Washing: Wash 7 times with a plate washer.

[0098] Primary antibody: Dilute the recombinant anti-PD-L1 humanized monoclonal antibody to 8800 ng / ml, with 11 concentrations in an arithmetic progression, mix it with 2×Tecentriq-Biotin in equal volume, add 100 μl / well to the enzyme-linked immunosorbent assay (ELISA) plate, and react at 37 °C for 1 hour.

[0099] Washing: Discard the supernatant and wash the plate 7 times with a plate washer.

[0100] Secondary antibody: Dilute Avidin-HRP 1:1000 according to the instruction manual, add 100 μl / well, and react at 37 °C for 1 hour.

[0101] Washing: Discard the supernatant and wash the plate 7 times with a plate washer.

[0102] Color development: Add the TMB substrate solution to the ELISA plate, 50 μl / well, and develop color in the dark for 15 minutes. Add the stop solution to the ELISA plate, 50 μl / well, and mix quickly.

[0103] Reading: Read the OD450nm with an ELISA reader, with 570nm as the reference wavelength.

[0104] Analysis of test results: Fit a curve with the sample concentration (abscissa, logarithmic value) against the OD value (ordinate), see Figure 7 , and calculate the EC50 value, as shown in the following table.

[0105] Table 12 Detection results of the competition binding of candidate humanized anti-PD-L1 monoclonal antibody with PD-1 to soluble PD-L1 Sample Name EC50 (ng / ml) Whether Competitive Tecentriq (Control) 2503 Competitive T0004-BC-LH 3598 Competitive T0004-C-LH 3294 Competitive

[0106] 1.2.8 Detection of biological activity of candidate humanized anti-PD-L1 monoclonal antibody Detection principle: The human T lymphoma cell line Jurkat stably expresses PD-1 on the cell membrane surface and the promoter of the NFAT-Luciferase reporter gene as effector cells, and Chinese hamster ovary cells (CHO-K1) stably express PD-L1 on the cell membrane surface as target cells.

[0107] In the absence of anti-PD-L1 monoclonal antibody, after the PD-1 on the surface of effector cells binds to PD-L1 on the surface of target cells, negative regulation occurs, and TCR activation will not be induced, so the luciferase reporter gene expression induced by the NFAT promoter will not be produced.

[0108] In the presence of anti-PD-L1 monoclonal antibody, the anti-PD-L1 monoclonal antibody can competitively bind to PD-L1 on the surface of target cells with PD-1 on the surface of effector cells. When target cells and effector cells are co-incubated, the TCR complex on the surface of Jurkat effector cells will be activated to different degrees according to the concentration of anti-PD-L1 monoclonal antibody (the higher the concentration, the higher the activation degree), thereby triggering different intensities of luciferase reporter gene expression induced by the NFAT promoter. The higher the antibody concentration, the stronger the luciferase reporter gene expression (the higher the RLU).

[0109] Preparation of target cells: Digest CHO-K1 / PD-L1 cells 1 - 2 days after subculture, and inoculate the cells into a 96-well white cell culture plate (do not let the cells grow to 100% confluence). Remove the culture medium, wash the cells with PBS, then add trypsin to digest the cells, and place the culture flask in a 37°C, 5% CO2 incubator for 3 - 5 minutes to make the cells round. Add 4 times the volume of cell culture medium to terminate digestion, and transfer the cell suspension to a 50 ml centrifuge tube. Take a small amount of cells, add trypan blue for counting, and centrifuge at 1000 rpm for 5 minutes. Discard the supernatant, and gently resuspend the cell pellet with target cell culture medium at 4×10 5 / ml, transfer the cell suspension to a sterile container, and add it to the corresponding wells of a 96-well cell culture plate at 100 μl / well (4×10 4 / well) (except for the BLK and single-effect wells). Add 100 μl / well of complete medium to the BLK and single-effect wells, and place the culture plate in a 37°C, 5% CO2 incubator overnight.

[0110] Dilution of antibody: On the day of detection, prepare an appropriate volume of detection culture medium freshly according to the number of samples.

[0111] Dilute the candidate humanized anti-PD-L1 monoclonal antibody with detection culture medium to 800 ng / ml (final concentration 400 ng / ml), and then continue to dilute it 1.5-fold with detection culture medium, for a total of 9 concentrations. The dilution factor for each step must be ≤10.

[0112] Preparation of effector cells: Collect Jurkat / NFAT-luc2 / PD-1 cells (Kyinno catalog number KC-1503) 2 days after subculture or amplification for inoculation. Take a small amount of cells, stain them with trypan blue for counting, and centrifuge at 1000 rpm for 5 minutes. Resuspend the pellet with detection culture medium at 1.25×10 6 / ml, and determine the resuspended volume according to the number of samples.

[0113] Inoculation of antibody and effector cells: Take out the 96-well cell culture plate that has been cultured overnight in the incubator. Use a pipette to discard 95 μl of the culture medium per well of the culture plate. Add the prepared antibody above. Add the diluted antibody to be detected at 40 μl per well. Among them, add 80 μl of the detection culture medium to the BLK wells and single-target wells, and add 40 μl of the detection culture medium to the single-effect and effect + target wells. Except for the BLK and single-target wells, add effector cells to the remaining wells at 40 μl per well (5×10 5 / well). Each 96-well cell culture plate should contain a detection culture medium control well (BLK), a single-target cell control well, a single-effect cell control well, and a target cell plus effector cell control well.

[0114] Continue to place the 96-well white cell culture plate in an incubator at 37°C and 5% CO2 for 6 hours and then perform detection.

[0115] Add 80 μl of the Bright-Glo Luciferase Assay System luminescent substrate per well and read the values using a GloMAX luminometer.

[0116] Analysis of the test results: Use the analysis software Origin to fit the standard curve: The abscissa (logarithm) is the concentration of the sample, and the ordinate is the luminescence value. The curve is in an "S" shape, as shown in Figure 8 . Select the four-parameter equation regression model. Calculate the half-maximal effective concentration EC50 according to the regression equation Y = (A1 - A2) / [1 + (X / x0)^p] + A2.

[0117] Table 13 Detection results of the bioactivity of the candidate humanized anti-PD-L1 monoclonal antibody Sample Name EC50 (ng / ml) T0004-BC-LH 55.22212 T0004-C-LH 56.86219 Tecentriq (Control) 53.09773

[0118] 1.2.9 Analysis of the ADCC effector function of the candidate humanized anti-PD-L1 monoclonal antibody Detection principle and method: Antibody-dependent cell-mediated cytotoxicity (ADCC) refers to the direct killing of antibody-coated target cells by cells expressing low-affinity Fc receptors (FcRⅢ) through recognition of the Fc segment of antibodies. CHO-K1 / hmPD-L1 cells express PD-L1 on the cell surface. Anti-PD-L1 monoclonal antibody binds to PD-L1 on the cell membrane surface. NK cells, macrophages, and neutrophils expressing FcR can directly kill target cells coated with IgG antibodies by binding to the Fc segment of the antibody. NK cells are the main effector cells mediating ADCC. The binding of the antibody to the antigen on the target cell is specific, while the killing effect of cells expressing FcR is non-specific. Our company has independently constructed a Jurkat NFAT-Re / CD16a(158V)-FcRγ / pcDNA3.1(hygro) reporter gene detection system, which can bind to the Fc segment of IgG1 antibody. If the effector cells are activated, its reporter gene will be initiated and expressed, promoting the luminescence of the luminescent substrate. By detecting the luminescence intensity, the degree of activation of the effector cells can be reflected, thereby reflecting the degree of activation of ADCC by the sample.

[0119] Target cell preparation: CHO-K1 / hmPD-L1 cells were passaged and cultured in 25 cm2 cell culture flasks or 10 cm cell culture dishes with 10% dialyzed FBS GMEM at 37 °C and 8% CO2, and used in this experiment during the logarithmic growth phase.

[0120] According to the experimental requirements, calculate the total number of cells needed, collect the cell suspension into a 15 ml centrifuge tube, centrifuge at 1000 rpm for 5 min, and discard the supernatant.

[0121] Wash once with ADCC analysis medium (4% FBS phenol red-free DMEM), centrifuge at 1000 rpm for 5 min, and discard the supernatant.

[0122] Continue to resuspend the cells with the above medium and adjust the cell density to 1×10 6 / ml.

[0123] Add the target cells to a 96-well clear-bottom white cell culture plate at 30 μl / well (except for BLK and single-effect wells). Add 30 μl / well of ADCC analysis medium to BLK and single-effect wells.

[0124] Incubation of target cells with antibody: Dilute anti-PD-L1 monoclonal antibody to 12000 ng / ml (final concentration 6000 ng / ml) with ADCC analysis medium. Serial dilute 3-fold with ADCC analysis medium for a total of 11 concentrations.

[0125] Add the diluted sample into a 96-well clear-bottom white cell culture plate as shown in the figure below, 60 μl per well. For BLK, single target, and single effector, add 60 μl per well of ADCC analysis medium, and for effector + target, add 30 μl per well of ADCC analysis medium.

[0126] Incubate in a 37 °C, 8% CO2 incubator for 30 - 60 minutes.

[0127] Preparation of effector cells: Jurkat NFAT-Re / CD16a(158V)-FcRγ / pcDNA3.1(hygro) cells are passaged and cultured in a 25 cm2 cell culture flask or a 10 cm cell culture dish using 10% FBS low-glucose RPMI1640 / DMEM medium. The culture conditions are 37 °C and 8% CO2. Cells in the logarithmic growth phase are used for this experiment.

[0128] Collect Jurkat NFAT-Re / CD16a(158V)-FcRγ / pcDNA3.1(hygro) cells in the logarithmic growth phase.

[0129] According to the experimental requirements, calculate the total number of cells needed. Collect the cell suspension into a 15 ml centrifuge tube, centrifuge at 1000 rpm for 5 min, and discard the supernatant.

[0130] Wash once with ADCC analysis medium, centrifuge at 1000 rpm for 5 min, and discard the supernatant.

[0131] Resuspend the cells with the above medium and adjust the cell density to 1×10 6 / ml.

[0132] Add the effector cells into the EC wells of the aforementioned 96-well clear-bottom white cell culture plate (except for the BLK and single target wells). Add 30 μl per well of ADCC analysis medium to the BLK and single target wells.

[0133] Incubate in a 37 °C, 8% CO2 cell culture incubator for 5 h.

[0134] Add 120 μl per well of Promega Bright-Glo ADCC luminescent substrate and read the values using a GloMAX luminometer.

[0135] Analysis of test results: Use the analysis software Origin to fit the standard curve: the abscissa (logarithmic) is the concentration of the sample, and the ordinate is the luminescence value. The curve is in an "S" shape, as shown in Figure 9 . Calculate the half-maximal inhibitory concentration EC50 according to the regression equation Y = (A1 - A2) / [1 + (X / x0)^p] + A2, as shown in the following table.

[0136] Table 14 Detection results of candidate humanized anti-PD-L1 monoclonal antibody ADCC Sample Name EC50 (ng / ml) T0004-BC-LH 6.39946 T0004-C-LH 7.00327 Tecentriq (Control) No ADCC Effect

[0137] 1.2.10 Analysis of CDC effector function of candidate humanized anti-PD-L1 monoclonal antibody Detection principle: Complement Dependent Cytotoxicity (CDC), that is, complement C1 can bind to the Fc segment of antibody molecules in the formed immune complex, thereby initiating the classical pathway of complement activation, and finally forming C5b-9 (membrane attack complex, MAC) on the surface of target cells. The small pores formed by MAC on the cell membrane lead to a decrease in intracellular osmotic pressure and cell lysis. In addition, the insertion of terminal complement components into the cell membrane may cause passive diffusion of lethal amounts of calcium ions into the cell, and ultimately lead to cell death. CHO-K1 / hmPD-L1 cells express PD-L1 on the cell surface. Recombinant humanized anti-PD-L1 monoclonal antibody can bind to PD-L1 on the cell surface, and the antibody hinge region can bind to complement C1q to observe whether the recombinant humanized anti-PD-L1 monoclonal antibody has CDC effector function.

[0138] Experimental method: Collect CHO-K1 / hmPD-L1 target cells in the logarithmic growth phase, wash them once with phenol red-free DMEM, and resuspend the precipitate at 3×10 5 / ml in phenol red-free DMEM, and add human serum as complement at 80 μl / ml, gently blow to mix evenly, and add 50 μl / well to a 96-well cell culture plate.

[0139] Dilute the recombinant humanized anti-PD-L1 monoclonal antibody with phenol red-free DMEM to 12000 ng / ml (final concentration 6000 ng / ml), and serially dilute it 3-fold with phenol red-free DMEM, for a total of 11 concentrations.

[0140] Add the diluted samples to a 96-well cell culture plate at 50 μl / well, and add 50 μl / well of phenol red-free DMEM to the BLK well.

[0141] Incubate in a 37 °C, 8% CO2 cell culture incubator for 4 h.

[0142] Prepare the chromogenic substrate according to the operation instructions of the Cytotoxicity Detection Kit (LDH). After mixing Cytotoxicity Detection Kit Plus Catalyst and Dye Solution at a volume ratio of 1:45. Add 100 μl / well of the mixed chromogenic substrate to the 96-well cell culture plate, develop color in the dark for 10 - 30 minutes (no need to cover the lid), and detect its OD490 with an enzyme-linked immunosorbent assay reader, with OD630 as the reference wavelength.

[0143] Analyze the half-maximal effective concentration (EC50) of the CDC effect of the recombinant anti-PD-L1 humanized monoclonal antibody.

[0144] Analysis of test results: Perform curve fitting on the sample concentration (abscissa, logarithmic value) against the OD value (ordinate), as shown in Figure 10 . Using the four-parameter regression model with the equation form Y = (A - D) / [1 + (X / C)^B] + D, calculate the half-maximal effective concentration (EC50), as shown in the following table.

[0145] Table 15 Detection results of candidate humanized anti-PD-L1 monoclonal antibody CDC Sample Name EC50 (ng / ml) T0004-BC-LH Weak CDC Effect at High Concentrations T0004-C-LH Weak CDC Effect at High Concentrations Tecentriq (Control) No CDC Effect

[0146] Analysis from 1.2.1 to 1.2.10 showed that both candidate humanized anti-PD-L1 monoclonal antibody regimens, T0004-BC-LH and T0004-C-LH, could bind to soluble PD-L1 and membrane-bound PD-L1, competitively bind to soluble PD-L1 and membrane-bound PD-L1 with the commercial antibody Tecentriq and PD-1, block the biological effects caused by the binding of PD-1 and PD-L1, and have ADCC effect and weak CDC effect. In the above series of analyses, the half-maximal effective dose EC50 of T0004-BC-LH was smaller than that of T0004-C-LH. Therefore, T0004-BC-LH was selected as the humanized regimen for the murine antibody E66 and entered further research and development as a candidate new drug. In subsequent studies, T0004-BC-LH was renamed "T0004". Example 2. Preparation of anti-PD-L1 antibody-drug conjugates (ADCs) with different DAR values

[0147] Use non-glycosylated Tecentriq (Atezolizumab, abbreviated as Ab1) and the humanized anti-human PD-L1 antibody T0004 (abbreviated as Ab2) prepared in Example 1, combined with a tumor microenvironment-responsive linker and the irinotecan derivative Dxd payload, to prepare tumor microenvironment-responsive anti-PD-L1 ADCs with different DAR values. The specific method steps are as follows.

[0148] 2.1 Conjugation of anti-PD-L1 monoclonal antibody with Dxd

[0149] Take 2 mg each of Ab1 and Ab2 and place them separately in 1.5 mL centrifuge tubes (n = 4 / group); add a 20-fold molar amount of tris(2-carboxyethyl)phosphine (TCEP, Sigma-Aldrich, CAS No. 51805-45-9) (diluted with 50 mM pH 7.2 phosphate buffer), react at 37 °C for 2 h to reduce the sulfhydryl groups of cysteine residues in the antibody; after supplementing 300 μL of phosphate buffer (PB) and 200 μL of dimethyl sulfoxide (DMSO), add the Dxd-Linker conjugate MC-GGFG-DXD (Deruxtecan, purchased from MedChemExpress, catalog number HY-13631E, CAS No. 1599440-13-7, chemical structural formula shown in Figure 22 )

[0150] Table 16 Antibody / Payload Molar Ratios Corresponding to Different DAR Values DAR Value DAR0 DAR2 DAR4 DAR6 DAR8 Antibody / Payload Molar Ratio 1:0 1:4 1:6 1:10 1:15

[0151] After coupling at 37 °C for 4 - 5 h, purify through a G25 desalting column (flow rate 0.5 mL / min) to obtain antibody-drug conjugates of Ab1 antibody and Ab2 antibody with DAR values of 0, 2, 4, 6, and 8, named Abs1-0, 2, 4, 6, 8 and Abs2-0, 2, 4, 6, 8 respectively. These antibody-drug conjugates have Figure 23 the structures shown

[0152] 2.2 Determination of DAR Value Preliminarily confirm the DAR value of the ADC by UV (ultraviolet method): Dilute the ADC concentration to 0.2 mg / mL, pipette 1 mL of the sample to be tested into a quartz cell with an optical path of 1 cm, set two absorption wavelengths of 280 nm and 370 nm, and detect with an Evolution300 ultraviolet spectrophotometer. Calculate the DAR value according to the formula

[0153] Confirm the DAR value of the ADC by SEC-HPLC (size exclusion chromatography) dual-wavelength detection method: Use an ACQUITY UPLC I-class BioAccord mass spectrometry system (purchased from Waters), select an ACQUITY SEC column (200 Å), the mobile phase is 200 mM PB (pH 6.8) containing 30% isopropanol, the flow rate is 0.2 mL / min, and dual-wavelength detection (280 / 370 nm).

[0154] Since antibodies have a maximum absorbance at 280 nm, while Dxd has a maximum absorbance at 370 nm in the ultraviolet range, by measuring the corresponding peak areas at the target wavelengths respectively and substituting them into the formula for calculation, the corresponding DAR value can be obtained. The calculation formula is as follows: DAR = (ε Ab370 - R * ε Ab280 ) / (R * ε D280 - ε D370 ) where R = A370 / A280, which is the ratio of the absorbance peak area of the ADC at 370 nm in the ultraviolet range to the absorbance peak area at 280 nm; ε Ab370 = 2042: the molar extinction coefficient of the antibody at 370 nm; ε Ab280 = 234900: the molar extinction coefficient of the antibody at 280 nm; ε D280 = 6650: the molar extinction coefficient of the drug Dxd at 280 nm; ε D370 = 19000: the molar extinction coefficient of the drug Dxd at 370 nm.

[0155] The DAR value measurement results of the series of ADCs prepared by the method in 2.1 are shown in the following table.

[0156] Table 17 DAR value measurement results

[0157] The experimental results show that the ADCs of both antibodies (Ab1 and Ab2) meet the expected DAR value gradient. This indicates that the preparation process can effectively control the uniformity and accuracy of the DAR value, providing a reliable basis for subsequent activity evaluation and stability analysis. Example 3. Detection of the physicochemical properties of anti-PD-L1 antibody conjugated drugs with different DAR values

[0158] 3.1 Polyacrylamide gel electrophoresis

[0159] Using the SDS-PAGE method (sodium dodecyl sulfate polyacrylamide gel electrophoresis), a 12% separating gel, Coomassie Brilliant Blue staining, and imaging analysis of the fragment distribution, the results are shown in Figure 11 .

[0160] The electrophoresis results showed that there was no obvious aggregation of the two antibodies after drug conjugation. Through the analysis of the electrophoresis bands, it was found that the sulfhydryl groups of the low-DAR ADCs were not completely occupied by Dxd. After removing the reducing agent, the free sulfhydryl groups were re-oxidized into disulfide bonds. Therefore, in non-reducing electrophoresis, as the DAR value decreased, the antibody disulfide bonds gradually recovered, and the bands of intact antibodies gradually increased with the decrease of the DAR value. At the same time, in reducing electrophoresis, the electrophoresis bands gradually increased from left to right, which was consistent with the increasing trend of the DAR value.

[0161] In addition, the electrophoresis results also showed that there was no obvious degradation of the ADC. This result further verified that our preparation process could effectively maintain the integrity and stability of the ADC.

[0162] 3.2 Isoelectric point (pI) analysis The CE Infinite whole-column imaging capillary isoelectric focusing electrophoresis system (iCIEF) was used to establish a pH gradient of 3 - 10 on the ciEF Cartridge. 50 μg of the sample was loaded, the pre-focusing voltage was 1500 V for 10 min, and the focusing voltage was 3000 V for 25 min. The results are shown in Figure 12 。

[0163] Through isoelectric focusing electrophoresis analysis, it was found that the isoelectric point (pI) of Ab2 was concentrated around 9.0, which was significantly higher than 8.7 of Ab1, indicating that there were differences in the surface charge distribution between the two. After conjugating with Dxd, the pI of Ab1 gradually decreased from 8.7 to the range of 8.1 - 8.5 as the DAR value increased, while the pI of the Ab2 conjugate only decreased slightly (8.9 when DAR = 8 and 9.1 when DAR = 2). This phenomenon suggested that the negatively charged groups of Dxd (such as carboxylic acid groups) could partially affect the pI of the ADC, but the degree of influence on different antibodies was different. It should be noted that when the antibody pI > 9, it is easy to have electrostatic adsorption with the negatively charged container surface, resulting in irreversible aggregation. Therefore, it is necessary to consider a suitable formulation buffer to preserve the ADC. Compared with the antibody, each DAR value ADC showed an increase in the pI heterogeneity characteristics to a certain extent, indicating that drug conjugation had a certain impact on the charge heterogeneity of the ADC. At the same time, there were also certain differences between high DAR (DAR = 6 / 8) and low DAR (DAR = 2 / 4), and there was a certain similarity between high DAR values, indicating that drug conjugation had a certain impact on charge heterogeneity, and with the saturation of conjugation, the charge heterogeneity tended to be stable.

[0164] 3.3 Purity analysis The purity analysis of ADC was performed by SEC-HPLC (size exclusion chromatography) using an ACQUITY UPLC I-class BioAccord mass spectrometry system (purchased from Waters Corporation). An ACQUITY SEC column (200 Å) was selected, and the mobile phase was 200 mM PB (pH 6.8) containing 30% isopropanol. The flow rate was 0.2 mL / min, and detection was performed at a wavelength of 280 nm. The analysis chromatogram is shown in Figure 13 .

[0165] The results showed that there was a small amount of aggregate in Abs2, and the content of dimer increased with the decrease of DAR value. There was no obvious dimer in ADCs with different DAR values of unglycosylated Ab1. This result was contrary to the common perception, indicating that the hydrophilicity / hydrophobicity of the antibody was not the determining factor affecting the aggregation stability of the ADC conjugation process. This finding provided a new idea for ADC process development. In addition to the optimization related to the linker and DAR value, the aggregation behavior of ADCs could also be regulated by optimizing the antibody.

[0166] 3.4 ADC Heterogeneity Analysis In the previous analysis, the two antibody ADCs showed different initial aggregate contents. Since papain treatment could reveal more hydrophobic characteristics of the domains, hydrophobic heterogeneity analysis was further performed on Abs1 / Abs2 with different DAR values by HIC-HPLC (hydrophobic interaction chromatography) after enzymatic digestion. The experimental method was as follows: The ADC was digested with papain. A TSKgel Phenyl-5PW column was selected, and gradient elution (1.5 M ammonium sulfate → 200 mM Tris-HCl + 30% isopropanol, 0.3 M ammonium sulfate) was performed at a flow rate of 1 mL / min for 20 min. The results are shown in Figure 14 .

[0167] The hydrophobic chromatography analysis showed that the overall peak elution time of Abs2 was earlier than that of Abs1, indicating that the overall hydrophobicity of Abs2 was less than that of Abs1. The main reason for the hydrophobicity difference between Abs2 and Abs1 might be the difference in the antibodies.

[0168] It is worth noting that Abs1 and Abs2 with different DAR values ​​showed different peak performances, especially as the DAR value increased, the proportion of high hydrophobic peaks of Abs2 gradually increased, indicating that the amount of coupled drug is another important factor affecting the hydrophobicity of PD-L1 ADC. There are also differences in the performance of hydrophobic characteristics between Abs1 and Abs2, and the proportion of peaks observable in the chromatographic separation area of ​​Abs1 does not show the same pattern as Abs2. Since the hydrophilicity of Ab2 antibody is better than that of Ab1, the ADC with a higher DAR value of Abs1 may be bound to the analytical column due to stronger hydrophobicity and cannot be effectively separated in the standard HIC analysis. There are many differences in amino acids in the variable regions of Ab1 and Ab2, and the hypervariable region of Ab2 contains more basic amino acids, which may be another factor for their difference in hydrophilicity. Example 4. Detection of functional activity of anti-PD-L1 antibody-drug conjugates with different DAR values

[0169] 4.1 ELISA to detect the binding activity of ADC to soluble PD-L1 Experimental methods: Recombinant human PD-L1 extracellular domain-Fc fusion protein was diluted to 0.01 μg / mL, 100 μL was coated in each well, and the plate was incubated in a water bath at 37 ℃ for 2 h; the plate was blocked overnight; the initial concentration of the sample to be tested was diluted to 5 μg / mL, and then 3-fold dilution was performed in 8 gradients, 100 μL was added to each well, 3 replicates were added, and the plate was incubated in a water bath at 37 ℃ for 2 h; the plate was washed 5-7 times; the goat anti-human HRP antibody was diluted 8000 times, 100 μL was added to the washed ELISA plate in each well, and the plate was incubated in a water bath at 37 ℃ for 2 h; 50 μL of the color development solution was incubated in the dark for 5 min, and 50 μL of the stop solution was used to stop the plate. The plate was scanned and read using a Molecular Devices ELISA reader at a scanning wavelength of 450 nm and a reference wavelength of 630 nm. The data were analyzed using Graph Pad software, see Figure 15 .

[0170] The results showed that regardless of the DAR value of 0, 2, 4, 6 or 8, the EC50 of Abs1 and Abs2 were stable at the sub-nM level, similar to the previous report (Atezolizumab-MMAE EC50 = 0.9 nM). This indicates that the introduction of tumor microenvironment-responsive linkers and Dxd payloads did not significantly change the CDR region conformation or epitope accessibility of the antibody, and the conjugated antibody had no significant effect on the affinity of PD-L1, and the two antibodies had similar performance. At the same time, the new Abs2 has similar binding activity to similar drugs, verifying the feasibility of Ab2 as an ADC carrier.

[0171] Notably, the ADC with a high DAR value still maintains a binding activity comparable to that of the naked antibody, suggesting that the payload attachment site is far from the antigen-binding domain, although the linker has been modified to improve hydrophilicity and tumor microenvironment release performance. This longer linker does not affect the antibody's binding domain.

[0172] 4.2 Detection of the cytotoxicity of ADC by the CCK-8 method To clarify the effect of the DAR value on the biological efficacy of ADC, in vitro cell killing experiments were performed using PD-L1-positive cells A431 and PD-L1-negative cells MC38. A431 cells are human epidermal cancer cells from ATCC (American Type Culture Collection), and MC38 cells are mouse colon cancer cells from ATCC. The expression of PD-L1 on the surface of the two types of cells was detected by fluorescence-labeled antibody combined with flow cytometry. The method was as follows: Cells were digested with 0.25% trypsin, washed 3-5 times with PBS, fluorescence-labeled anti-PD-L1 antibody was added to the cell suspension, incubated for 30 minutes, washed 3-5 times with PBS, and detected by flow cytometry. Data analysis was performed using FACSDiva Software. The results are shown in Figure 16 , verifying that A431 cells are PD-L1 positive and MC38 cells are PD-L1 negative.

[0173] The Cell Counting Kit-8 kit (purchased from MedChemExpress, product catalog number HY-K0301, abbreviated as CCK-8), which is a detection reagent based on WST-8 (water-soluble tetrazolium salt-8) and is widely used in the detection of cell proliferation and cytotoxicity. The survival rate of A431 cells / MC38 cells was measured according to the method in the kit instructions (72 h treatment, n = 3 replicates). The detection wavelength of the Molecular Devices microplate reader was 450 nm, and the reference wavelength was 600 nm for reading. Data analysis was performed using Graphpad software. The results are shown in Figure 17 and Figure 18 .

[0174] The detection results showed that in PD-L1-positive cells, the cytotoxicity of Abs1 and Abs2 both increased significantly with the increase of the DAR value, and the cell killing activity was enhanced, which was in line with the expectation that a high drug load can improve the killing efficiency in the "bystander effect": Abs1 EC50: 0.4 μM when DAR = 8, 1.2 μM when DAR = 2; Abs2 EC50: 0.3 μM when DAR = 8, 1.0 μM when DAR = 2, Neither of the two ADCs showed significant cytotoxicity against PD-L1-negative cells, confirming that their toxicity depends on PD-L1-mediated endocytosis and tumor microenvironment-responsive drug release. Meanwhile, with DAR values between 6 and 8, the cytotoxicity results were similar, indicating that the relationship between DAR value and activity is not a simple linear one. Each ADC needs to form an optimal DAR to ensure the unity of safety, effectiveness, and drug-likeness characteristics. Abs2 had a similar or slightly stronger killing efficacy at DAR 8 (EC50 = 0.3 μM) compared to Abs1 (0.4 μM), which may be related to antibody-related differences such as binding epitopes, Fc receptor function, and even non-specific killing, etc.

[0175] Compared with free Dxd (EC50 = 6.9 μM), the cytotoxicity of the ADC was significantly improved ( Figure 18 ), and this synergistic effect was significantly higher than that of traditional ADC drugs, highlighting the synergistic advantage of the tumor microenvironment-responsive linker and the highly active payload. Example 5. Evaluation of the stability of anti-PD-L1 antibody-drug conjugates with different DAR values

[0176] ADCs have higher chemical instability compared to classical antibody drugs, and the stability of ADCs is one of the core indicators of their drug-likeness. Especially for ADCs targeting immune checkpoints, it is necessary to balance the activity advantage brought by high DAR values and the potential stability risks. Therefore, it is crucial to conduct a preliminary stability assessment of ADCs at the early stage of drug development.

[0177] 5.1 Thermal stability analysis of ADCs Differential scanning calorimetry (DSC) is an important tool for evaluating the thermal stability of antibody drugs. The DSC analysis of Abs1 and Abs2 with different DAR values was carried out using a MicroCal VP-Capillary differential scanning calorimeter (purchased from Malvern). The results are shown in the following table.

[0178] Table 18 Thermal stability analysis of PD-L1 ADCs with different DAR values ADC Tm onset ℃ Tm 1 ℃ Abs1-0 76.14 84.64 Abs1-2 54.11 82.04 Abs1-4 55.36 81.06 Abs1-6 56.48 80.40 Abs1-8 60.43 80.31 Abs2-0 66.71 79.00 Abs2-2 55.11 65.40 Abs2-4 56.82 65.54 Abs2-6 56.53 78.00 Abs2-8 56.93 77.71

[0179] The results showed that, compared with the control (DAR = 0), the Tm onset (onset melting temperature) and Tm1 (first melting peak temperature) of PD-L1 ADCs with different DAR values were significantly decreased, indicating that drug conjugation caused certain perturbations to the native conformation of the antibody, and the conjugation of the linker alone could present such perturbations. However, the effect of DAR value on thermal stability showed a non-linear characteristic. There were certain differences in Tm onset or Tm1 between lower DAR values and higher DAR values, suggesting that the heterogeneity of conjugation might also exacerbate conformational instability at low drug-loading levels. There were also differences in thermal stability between Abs1 and Abs2. Specifically, Abs1 and Abs2 showed similar Tm onset, but the Tm1 of Abs1 was slightly higher, and the high-DAR ADCs had slightly higher Tm onset or Tm1. These results broke through the traditional understanding that lower DAR values are more stable, indicating that moderately high DAR values might compensate for conjugation damage by improving structural homogeneity.

[0180] In the analysis of Abs1 and Abs2 with different DAR values, the high-DAR ADCs did not show a trend of decreased thermal stability, and the two ADCs showed similarity in thermal stability. This conjugation strategy and linker could support the preparation of PD-L1 ADCs with higher DAR values without having a significant negative impact on stability.

[0181] 5.2 Preliminary stability assessment of ADC

[0182] The anti-PD-L1 ADCs with different DAR values were subjected to accelerated stability assessment in a weakly acidic buffer at 2 - 8 °C. Specific method: Each DAR value of ADC was diluted to a concentration of 1 mg / mL with acetic acid buffer at pH 5.8 (containing 0.137 mmol / mL acetic acid, 0.02 mmol / mL L-histidine, 0.4 mg / mL polysorbate 20, 0.12 mmol / mL sucrose), and 4 replicates were prepared. They were stored in the dark at 2 - 8 °C, and samples were taken for testing on the 0th day (0D), 20th day (20D), 40th day (40D), and 60th day (60D) respectively.

[0183] SDS-PAGE analysis was carried out according to the method of Example 3.1, and the results are shown in Figure 19 . The results showed that within 60 days of storage at 2 - 8 °C, the anti-PD-L1 ADCs with different DAR values did not show obvious degradation.

[0184] SEC analysis was further carried out according to the method of Example 3.3, and the results are shown in Figure 20The results showed that with the increase of storage time, SEC detection was carried out on ADCs with different DAR values at four different times respectively. The proportion of aggregates of both ADCs remained at about 5% or less. Although the initial dimer of Abs2 was higher than that of Abs1, the dimer growth of Abs2 with high DAR value was significantly slower than that of Abs2 with low DAR value, which was consistent with the results of DSC thermal stability analysis, clarifying the advantage of Ab2 as a novel PD-L1 antibody in supporting the preparation of novel PD-L1 ADC, and both ADCs with high DAR values showed good stability in the stability evaluation.

[0185] To verify the stability of the ADC linker, we further detected the DAR value by SEC dual wavelength using the method of Example 2.2. The results showed that the change of DAR value was not obvious, especially for the ADC with DAR = 6 (see the data table in Example 2.2).

[0186] To verify that it would not affect the binding activity of the antibody during storage, we also detected the ADCs at four time points by ELISA using the method in Example 4.1. The results showed that the ADCs with different DAR values stored in liquid state at 2 - 8 °C for two months did not affect their binding activity. See Figure 21 .

[0187] Through the preliminary stability evaluation, we found that within two months of storage at 2 - 8 °C in the unoptimized basic buffer solution, the basic functions of ADCs with various DAR values remained stable. Example 6. Study on the effect of anti-PD-L1 antibody conjugated drugs with different DAR values on canine tumor cells

[0188] Compared with classical antibody drugs, ADCs have higher chemical instability, and the stability of ADCs is one of the core indicators of their druggability. Especially for ADCs targeting immune checkpoints, it is necessary to balance the activity advantages brought by high DAR values and potential stability risks. Therefore, it is crucial to conduct a preliminary stability evaluation of ADCs in the early stage of drug development.

[0189] 1.6.2 Antibody and ADC function evaluation

[0190] Based on the canine PD-L1 sequence in the NCBI database (GenBank: accession number AB898678.1), the pL101 expression vector was constructed by codon optimization. The vector was constructed with HindIII / NheI double digestion sites and fused with the mouse Fc fragment for expression. After transfection with the ExpiCHO-Sh transient expression system (which retained the post-translational glycosylation modification of the PD-L1 protein) for 14 days, it was purified by MabCapture-R protein A affinity chromatography, and finally the soluble canine PD-L1-Fc fusion protein was obtained.

[0191] 1.6.2 Antibody and ADC Function Evaluation.

[0192] 1.6.2.1 Detection of the binding ability to soluble canine PD-L1 by enzyme-linked immunosorbent assay: Coat the enzyme-linked immunosorbent assay plate with canine PD-L1 protein (10 μg / mL, prepared in step 1.4.2) using pH 9.6 carbonate buffer at 37 °C for 1.5 h. After blocking, successively add serially diluted antibody (ADC) (initial concentration 50 μg / mL, 2-fold serial dilution, 10 dilutions), HRP-labeled secondary antibody, and chromogenic solution. Measure the absorbance at 450 / 630 nm using a SpectraMax M5 microplate reader (Molecular Devices). Data is collected by SoftMax Pro 6.3 and analyzed by GraphPad Prism. The measurement results are shown in Figure 24 , and the results show that Abs1 and Abs2 exhibit concentration-dependent binding to canine PD-L1.

[0193] 1.6.2.2 Detection of the affinity for canine PD-L1 protein by surface plasmon resonance (SPR): Use a Biacore T200 surface plasmon resonance instrument (GE Healthcare). Immobilize 500 RU of canine PD-L1 protein on a CM5 sensor chip. Using HBS-EP as the buffer, perform kinetic analysis in multi-cycle (for monoclonal antibody A) and single-cycle (for monoclonal antibody T) modes according to different affinities. The KD value is obtained by fitting the data with Biacore T200 Evaluation Software. The results are shown in the following table. The affinity of Abs1 for canine PD-L1 can reach the nM level, while the affinity of Abs2 for canine PD-L1 is approximately in the μM level. The coupling has no obvious effect on the binding activity to canine PD-L1.

[0194] Table 19 Affinity of ADCs with different DAR values for soluble canine PD-L1 determined by SPR ADC Number Affinity (M) Abs1-0 2.79E-10 Abs1-2 3.214E-10 Abs1-4 2.860E-10 Abs1-6 3.033E-10 Abs1-8 3.150E-10 Abs2-0 5.068E-5 Abs2-2 1.813E-7 Abs2-4 3.864E-7 Abs2-6 2.138E-7 Abs2-8 2.087E-7

[0195] 1.6.2.3 Detection of the binding ability to cell membrane surface canine PD-L1 by flow cytometry: Use canine tumor cells (canine mammary tumor cell PETCC2478, canine head dermatofibrosarcoma protuberans cell PETCC2477, provided by the public welfare organization PETCC, the Canine and Feline Cell Line Preservation Center) at 1.5×10 6 / mL, co-incubated with antibodies at gradient concentrations (45 min, protected from light), washed, and then FITC-labeled secondary antibody was added (45 min, ice bath). Detection was performed using a FACS Canto II flow cytometer (BD), and analysis was carried out using FACSDiva Software. The results are shown in Figure 25 . The results showed that for PD-L1 on the canine cell membrane, both Ab1 and Ab2 could bind saturably at certain concentrations. The difference in the binding ability on the cell membrane surface and the results of solution-phase detection might be due to: (1) conformational heterogeneity of PD-L1 on the cell membrane surface; (2) recognition preference of T monoclonal antibody for glycosylated epitopes; (3) regulatory effect of accessory proteins in the tumor microenvironment on epitope exposure.

[0196] 1.6.2.4 Cytotoxicity experiment:

[0197] The above-mentioned canine tumor cells (5×10 3 / well) were co-cultured with ADC at gradient concentrations for 72 h. The absorbance at 450 nm was measured using a Cell Counting Kit-8 (purchased from MedChemExpress, product catalog number HY-K0301, abbreviated as CCK-8), and the IC50 value was calculated using GraphPad Prism. The results are shown in Figure 26 、 Figure 27 .

[0198] The results showed that both Abs1 and Abs2 types of ADC had a killing effect on canine tumor cells, and the killing efficacy showed an obvious DAR value dependence: the killing efficacy of DAR 6 - 8 ADC on PETCC2478 cells was significantly improved compared with that of DAR2 - 4. Abs1 and Abs2 antibodies showed similar DAR value-dependent performance. Control experiments confirmed that irrelevant antibody ADC and low-free Dxd did not show significant cytotoxicity, demonstrating that the killing effect was target-specific. Abs2-6 and Abs2-8 showed a higher cell inhibition rate on PETCC2478 cells than the Abs1 series of ADC ( Figure 26 ).

[0199] It should be noted that although the solution-phase affinity of Ab2 monoclonal antibody and canine PD-L1 was lower than that of Ab1, Abs2-8 still showed a killing effect comparable to that of Abs1-type ADC. This might be attributed to: (1) antigen cross-linking effect; (2) free Dxd released by linker cleavage killed adjacent low-antigen cells through the bystander effect; (3) multivalent binding on the membrane surface compensated for the insufficient affinity at a single site; (4) different endocytic characteristics might compensate for the insufficient affinity.

Claims

1. An anti-PD-L1 antibody conjugate drug, the molecule of which is formed by conjugating an anti-PD-L1 antibody and a payload through a linker, characterized in that, The variable region sequence of the antibody light chain is the sequence recited in SEQ ID NO:1, and the variable region sequence of the heavy chain is the sequence recited in SEQ ID NO:

2.

2. The anti-PD-L1 antibody conjugate according to claim 1, characterized in that, The Fc segment of the antibody is of the IgG1 type.

3. The anti-PD-L1 antibody conjugate according to claim 1, wherein The light chain sequence of the antibody is the sequence recited in SEQID NO:3, and the heavy chain sequence is the sequence recited in SEQ ID NO:

4.

4. The anti-PD-L1 antibody conjugate drug according to claim 1, wherein The ratio of the payload to the antibody (DAR) is 6 to 8 (inclusive).

5. The anti-PD-L1 antibody conjugate according to claim 1, characterized in that, The ratio of the payload to the antibody (DAR) is 8.

6. The anti-PD-L1 antibody conjugate according to claim 1, wherein, The payload is the irinotecan derivative Dxd.

7. The anti-PD-L1 antibody conjugate according to claim 1, wherein, The payload is conjugated to the anti-PD-L1 antibody through a cysteine residue.

8. The anti-PD-L1 antibody conjugate according to claim 1, wherein The linker is maleimide-GGFG peptide.

9. A method for preparing the anti-PD-L1 antibody conjugate drug as claimed in claim 1, characterized in that, Comprising the following steps: Mix the anti-PD-L1 antibody recited in claim 1 with 20-fold molar amount of TCEP (tris(2-carboxyethyl)phosphine), and reduce at 37 °C for 2 hours; Add the Dxd-Linker conjugate according to the antibody:drug molar ratio of 1:10 to 1:15, and couple at 37 °C for 4-5 hours; Desalt and purify to obtain the anti-PD-L1 antibody conjugated drug.

10. The method for preparing an anti-PD-L1 antibody conjugate according to claim 8, wherein The Dxd-Linker conjugate is MC-GGFG-DXD (Deruxtecan).

11. The method for preparing an anti-PD-L1 antibody conjugate drug according to claim 8, wherein, The anti-PD-L1 antibody is dissolved in phosphate buffer at pH 7.2 at a concentration of 10 mg / ml, and the Dxd-Linker conjugate is dissolved in anhydrous DMSO at a concentration of 10 mmol / L.

12. Use of the anti-PD-L1 antibody conjugated drug recited in claim 1 in the preparation of a canine tumor drug.

Images