Device, method and computer-readable storage medium for designing humanized antibodies based on CDR transplantation

Abstract

The present invention discloses an apparatus, method, and computer-readable storage medium for designing humanized antibodies based on CDR transplantation. The method of the present invention obtains a CDR region-transplanted antibody by replacing the antigen complementary determining region sequence of a humanized antibody sequence with the antigen complementary determining region sequence of a heterologous antibody to a target antigen; performs single amino acid back mutation on the different amino acid sites between the CDR region-transplanted antibody and the heterologous antibody, and compares the structural difference scores of the three-dimensional structure of the protein before and after the mutation; retains the single amino acid back mutation sites with a structural difference score of less than or equal to 1, and finally obtains a CDR-transplanted humanized antibody for the target antigen. The method of the present invention can increase the similarity between the heterologous antibody and the human antibody sequence, reduce the immunogenicity of the human antibody to the target antigen, and maintain affinity.

Description

Technical Field

[0001] The present invention belongs to the technical field of bioinformatics, and in particular relates to a device, method and computer-readable storage medium for designing humanized antibodies based on CDR transplantation. Background Art

[0002] Injection of heterologous antibodies into the human body can cause ADA. Common approaches to address this issue include chimeric antibodies, CDR grafting, SDR transplantation, fully human immunization, and immune site extraction.

[0003] Modified antibodies, also known as CDR-grafted antibodies, are the regions of the antibody variable region (CDRs) that recognize and bind to antigens and directly determine the antibody's specificity. By transplanting the CDRs of a heterologous monoclonal antibody into the variable region of a human antibody, replacing the human antibody CDRs, the human antibody acquires the antigen-binding specificity of the heterologous monoclonal antibody while reducing its heterologous nature. However, while the antigen primarily contacts the antibody's CDRs, the FR region often also participates, influencing the spatial configuration of the CDRs. Therefore, after replacing the human FR region, the V region, where the heterologous CDRs and human FRs are embedded, may alter the original CDR configuration of the monoclonal antibody, resulting in a decrease or even a significant reduction in antigen binding ability. Although molecular design of antibodies is now possible, introducing certain key amino acid residues from the heterologous FR region into the human FR region can, if properly configured, achieve an affinity comparable to that of the original mouse antibody. However, humanized antibodies often fail to achieve the affinity of the original heterologous monoclonal antibody.

[0004] A common method for humanizing non-humanized antibodies is complementarity-determining region (CDR) grafting, where the CDR regions of a non-human antibody are transplanted onto the framework regions of a human antibody. Typically, the human antibody framework region with the highest homology to the non-human antibody framework region is selected as the recipient for the CDR grafting. The primary drawback of this approach is that affinity for binding to a specific target can be reduced or even lost. Directly transplanting the CDR loops of a mouse antibody onto the human antibody framework may not affect antibody affinity in some cases, but in many cases it can significantly reduce affinity. Some residues in the framework regions of mouse antibodies, known as vernier region residues, have been shown to affect the conformation of the CDR loops and the affinity of the antibody. These residues are located near the β-sheets of the CDR regions. Therefore, after selecting the desired human antibody framework regions, these residues need to be backmutated to retain them in the humanized antibody. Furthermore, mutations in amino acid residues outside the variable regions have also been used to impart new properties to humanized antibodies. However, finding suitable backmutation sites for these residues is difficult with existing technologies.

[0005] The concept of the B-factor, a temperature factor, originated in crystallography. It primarily reflects the ambiguity of atomic conformational states within a crystal, effectively reflecting the conformational state of a protein molecule within the crystal. A higher B-factor indicates greater ambiguity, indicating a more unstable or less stable conformation at the corresponding site. In crystallographic data, the B-factor is typically given in atomic units and is often converted to the B-factor value of the corresponding amino acid residue, allowing analysis of the conformational stability or flexibility of that residue. Studies have shown that mutating amino acids with high B-factor values ​​in protein structures can significantly improve the thermal stability of some mutants. Summary of the Invention

[0006] The technical problem to be solved by the present invention is how to transform heterologous antibodies to obtain humanized antibodies and / or how to reduce the immunogenicity of heterologous antibodies and / or how to reduce the immunogenicity of heterologous antibodies while maintaining their affinity to antigens.

[0007] In order to solve the above technical problems, the present invention first provides a device for obtaining (designing) humanized antibodies based on CDR region transplantation, which may include the following modules:

[0008] A1) Humanized framework region selection module: used to obtain humanized antibody sequences of heterologous antibodies specific for target antigens using sequence alignment software;

[0009] A2) CDR region transplantation module: used to annotate the antigen complementary determining region sequence of the heterologous antibody using an antibody sequence CDR annotation system to obtain a heterologous antibody CDR sequence, and annotate the humanized sequence to obtain a humanized antibody CDR sequence; and replace the humanized antibody CDR sequence in the humanized antibody sequence with the heterologous antibody CDR sequence to obtain a CDR region transplanted antibody sequence for the target antigen;

[0010] A3) a module for selecting back mutation sites and obtaining humanized antibodies: used to align the sequence of the CDR region-grafted antibody with the sequence of the heterologous antibody to obtain differential amino acid sites, perform back mutation on each of the differential amino acid sites to obtain single amino acid back mutation CDR region-grafted antibodies, perform structural prediction on all of the single amino acid back mutation CDR region-grafted antibodies using protein structure prediction software to obtain the three-dimensional structures of all of the single amino acid back mutation CDR region-grafted antibodies, and use the protein structure prediction software to predict the three-dimensional structure of the heterologous antibody CDR sequence to obtain the three-dimensional structure of the heterologous antibody;

[0011] The three-dimensional structure of the single amino acid back-mutated CDR region transplanted antibody and the three-dimensional structure of the heterologous antibody are scored for structural difference using the following formula 1 to obtain a structural difference score;

[0012] score=important_score×exposed_area×3D_smiliarity Formula 1;

[0013] In Formula 1, score represents the structural difference score between the three-dimensional structure of the single amino acid reverted CDR region grafted antibody and the three-dimensional structure of the heterologous antibody; important_score is the value obtained by scoring the similarity of the three-dimensional structure of the single amino acid reverted CDR region grafted antibody and the three-dimensional structure of the heterologous antibody based on the B-factor and hydrophilicity of each amino acid residue in the sequence; exposed_area represents the difference in the size of the exposed area of ​​the three-dimensional structure of the heterologous antibody and the three-dimensional structure of the single amino acid reverted CDR region grafted antibody; 3D_similarity is the structural similarity score between the three-dimensional structure of the single amino acid reverted CDR region grafted antibody and the three-dimensional structure of the heterologous antibody (representing the similarity of the protein three-dimensional structure before and after the single amino acid site reverted mutation);

[0014] The amino acid mutation sites in the single amino acid back mutation CDR region transplanted antibody with a structural difference score less than 1 are selected and merged to obtain back mutation combination sites, and the amino acid residues at the back mutation combination site position in the sequence of the CDR region transplanted antibody are replaced with the amino acid residues at the back mutation combination site position corresponding to the heterologous antibody to obtain a humanized antibody sequence for the target antigen.

[0015] The complementary determining region sequences include sequences of the CDR1, CDR2 and CDR3 regions.

[0016] In the above device, the heterologous antibody may be a non-human antibody.

[0017] In the above device, in formula 1 of A3), the important_score can be calculated using the abnumber program package.

[0018] In formula 1 of A3), the exposed_area can be calculated using the colabfold program package. In formula 1 of A3), the 3D_similarity can be calculated using the pymol program package.

[0019] In the above device, the antibody sequence CDR annotation system is a CDR annotation system that combines Kabat and Chothia numbering.

[0020] In order to solve the above technical problems, the present invention also provides a method for obtaining (designing) a humanized antibody based on CDR region transplantation, which may include the following steps:

[0021] B1) Selection of humanized framework regions: Use sequence alignment software to obtain the humanized antibody sequence of the target antigen heterologous antibody;

[0022] B2) CDR region transplantation: using an antibody sequence CDR annotation system to annotate the antigen complementary determining region sequence of the heterologous antibody to obtain a heterologous antibody CDR sequence, and annotating the humanized antibody sequence to obtain a humanized antibody CDR sequence; replacing the humanized antibody CDR sequence in the humanized antibody sequence with the heterologous antibody CDR sequence to obtain a CDR region transplanted antibody sequence for the target antigen;

[0023] B3) Selection of back mutation sites and obtaining humanized antibodies: comparing the sequence of the CDR region-grafted antibody with the sequence of the heterologous antibody to obtain differential amino acid sites, performing back mutation on each of the differential amino acid sites to obtain single amino acid back mutation CDR region-grafted antibodies, performing structural prediction on all of the single amino acid back mutation CDR region-grafted antibodies using protein structure prediction software to obtain the three-dimensional structures of all of the single amino acid back mutation CDR region-grafted antibodies, and using the protein structure prediction software to predict the three-dimensional structure of the heterologous antibody CDR sequence to obtain the three-dimensional structure of the heterologous antibody;

[0024] The three-dimensional structure of the single amino acid back-mutated CDR region transplanted antibody and the three-dimensional structure of the heterologous antibody are scored for structural difference using the following formula 1 to obtain a structural difference score;

[0025] score=important_score×exposed_area×3D_smiliarity Formula 1;

[0026] In Formula 1, score represents the structural difference score between the three-dimensional structure of the single amino acid reverted CDR region grafted antibody and the three-dimensional structure of the heterologous antibody; important_score is the value obtained by scoring the similarity of the three-dimensional structure of the single amino acid reverted CDR region grafted antibody and the three-dimensional structure of the heterologous antibody based on the B-factor and hydrophilicity of each amino acid residue in the sequence; exposed_area represents the difference in the size of the exposed area of ​​the three-dimensional structure of the heterologous antibody and the three-dimensional structure of the single amino acid reverted CDR region grafted antibody; 3D_similarity is the structural similarity score between the three-dimensional structure of the single amino acid reverted CDR region grafted antibody and the three-dimensional structure of the heterologous antibody (representing the similarity of the protein three-dimensional structure before and after the single amino acid site reverted mutation);

[0027] The amino acid mutation sites in the single amino acid back mutation CDR region transplanted antibody with a structural difference score less than 1 are selected and merged to obtain back mutation combination sites, and the amino acid residues at the back mutation combination site position in the sequence of the CDR region transplanted antibody are replaced with the amino acid residues at the back mutation combination site position corresponding to the heterologous antibody to obtain a humanized antibody sequence for the target antigen.

[0028] The heterologous antibodies mentioned above are non-human antibodies.

[0029] The complementary determining region sequences mentioned above include sequences of the CDR1, CDR2 and CDR3 regions.

[0030] In the above method, the antibody sequence CDR annotation system can be a CDR annotation system that combines Kabat and Chothia numbering.

[0031] In the above method, in formula 1 of B3), the important_score can be calculated using the abnumber program package; in formula 1 of B3), the exposed_area can be calculated using the colabfold program package; in formula 1 of B3), the 3D_similarity can be calculated using the pymol program package.

[0032] In order to solve the above technical problem, the present invention further provides a computer-readable storage medium storing a computer program, wherein the computer program can enable a computer to execute the steps of the method described above.

[0033] In order to solve the above technical problem, the present invention further provides a computer device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor can execute the computer program to implement the following steps:

[0034] B1) Selection of humanized framework regions: Use sequence alignment software to obtain the humanized antibody sequence of the target antigen heterologous antibody;

[0035] B2) CDR region transplantation: using an antibody sequence CDR annotation system to annotate the antigen complementary determining region sequence of the heterologous antibody to obtain a heterologous antibody CDR sequence, and annotating the humanized antibody sequence to obtain a humanized antibody CDR sequence; replacing the humanized antibody CDR sequence in the humanized antibody sequence with the heterologous antibody CDR sequence to obtain a CDR region transplanted antibody sequence for the target antigen;

[0036] B3) Selection of back mutation sites and obtaining humanized antibodies: comparing the sequence of the CDR region-grafted antibody with the sequence of the heterologous antibody to obtain differential amino acid sites, performing back mutation on each of the differential amino acid sites to obtain single amino acid back mutation CDR region-grafted antibodies, performing structural prediction on all of the single amino acid back mutation CDR region-grafted antibodies using protein structure prediction software to obtain the three-dimensional structures of all of the single amino acid back mutation CDR region-grafted antibodies, and using the protein structure prediction software to predict the three-dimensional structure of the heterologous antibody CDR sequence to obtain the three-dimensional structure of the heterologous antibody;

[0037] The three-dimensional structure of the single amino acid back-mutated CDR region transplanted antibody and the three-dimensional structure of the heterologous antibody are scored for structural difference using the following formula 1 to obtain a structural difference score;

[0038] score=important_score×exposed_area×3D_smiliarity Formula 1;

[0039] In Formula 1, score represents the structural difference score between the three-dimensional structure of the single amino acid reverted mutation CDR region grafted antibody and the three-dimensional structure of the heterologous antibody; important_score is the value obtained by scoring the similarity of the three-dimensional structure of the single amino acid reverted mutation CDR region grafted antibody and the three-dimensional structure of the heterologous antibody based on the B-factor and hydrophilicity of each amino acid residue in the sequence; exposed_area represents the difference in the size of the exposed area of ​​the three-dimensional structure of the heterologous antibody and the three-dimensional structure of the single amino acid reverted mutation CDR region grafted antibody; 3D_similarity is the structural similarity score between the three-dimensional structure of the single amino acid reverted mutation CDR region grafted antibody and the three-dimensional structure of the heterologous antibody;

[0040] The amino acid mutation sites in the single amino acid back mutation CDR region transplanted antibody with a structural difference score less than 1 are selected and merged to obtain back mutation combination sites, and the amino acid residues at the back mutation combination site position in the sequence of the CDR region transplanted antibody are replaced with the amino acid residues at the back mutation combination site position corresponding to the heterologous antibody to obtain a humanized antibody sequence for the target antigen.

[0041] In order to solve the above technical problem, the present invention also provides a computer device, including a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the above method.

[0042] The heterologous source mentioned above may be non-human.

[0043] The present invention increases the similarity between heterologous antibodies and human antibody sequences through CDR grafting, thereby reducing immunogenicity and maintaining affinity. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic flow chart of the present invention.

[0045] Figure 2 These are the Elisa EC50 test results of the CDR-grafted protein sequence thera_52 and the parent CD28 antibody after site back mutation in Example 1, respectively, with CD28.

[0046] Figure 3 These are the affinity test results of hu_PD-L1 and PD-L1 antibody parent with PD-L1 in Example 2. DETAILED DESCRIPTION

[0047] The present invention will be further described in detail below in conjunction with specific embodiments. The examples provided are only for illustrating the present invention and are not intended to limit the scope of the present invention. The examples provided below can serve as a guide for further improvements by those skilled in the art and are not intended to limit the present invention in any way.

[0048] Unless otherwise specified, the experimental methods in the following examples are conventional methods and were performed according to the techniques or conditions described in the literature in the field or according to the product instructions. The materials and reagents used in the following examples, unless otherwise specified, were all commercially available.

[0049] The antibody expression in the embodiment of the present invention is carried out by genetic engineering technology, using the protein synthesis system of eukaryotic or prokaryotic host cells to synthesize the corresponding antibody protein, and the method is as follows:

[0050] Primer design: Design and synthesize specific primers for the antibody gene and add appropriate restriction endonuclease recognition sequences to the 5' end of the primers.

[0051] PCR amplification: Using designed primers and template DNA, antibody gene fragments are amplified by PCR technology.

[0052] Enzyme digestion reaction: The PCR product and cloning vector are digested with the same restriction endonuclease to produce sticky ends.

[0053] Ligation reaction: connect the digested antibody gene fragment to the vector using DNA ligase.

[0054] Transformation: Transform the ligation product into competent E. coli.

[0055] Screening and identification: Positive clones were screened by antibiotics, and colony PCR and sequence analysis were performed to confirm the correctness of the inserted fragment.

[0056] Expression and purification: The verified correct plasmid is transformed into a eukaryotic or prokaryotic expression system to induce the expression of the antibody protein, and the protein is purified by affinity chromatography and other methods.

[0057] The method for antibody affinity testing in the embodiment of the present invention is as follows:

[0058] The antibody affinity was detected by SPR using a biomolecular interaction instrument (Biacore 8K).

[0059] 1) Chip activation: NHS (N-hydroxysuccinimide) was used to activate the chip surface.

[0060] 2) Antibody immobilization: The purified target antibody protein is diluted to an appropriate concentration and injected into the activated chip channel to be fixed on the surface.

[0061] 3) Blocking: A high concentration of blocking agent (bovine serum albumin) is injected into the chip to reduce nonspecific binding.

[0062] 4) Baseline correction: Ensure that the baseline of the chip is stable before injecting the antigen.

[0063] 5) Antigen injection:

[0064] Prepare a series of antigen solutions with different concentrations and inject them sequentially into the channel containing the antibody. Record the response units (RU) during the binding and dissociation processes.

[0065] 6) Data Acquisition and Analysis: After data collection, perform kinetic analysis using BIACORE software. Fit the data using a 1:1 Langmuir binding model or other appropriate model. Calculate the KD value, or dissociation constant, a key parameter for measuring affinity.

[0066] The ELISA method for detecting the antigen binding activity of antibodies in the embodiment of the present invention is as follows:

[0067] Antigen preparation: Prepare a series of antigen solutions at different concentrations (10 concentrations) to cover the expected EC50 values.

[0068] Antibody preparation: Prepare antibodies of known concentration.

[0069] ELISA plate: Use a 96-well ELISA plate.

[0070] Reagents: Prepare enzyme-labeled secondary antibody (horseradish peroxidase-labeled anti-antibody), substrate solution, stop solution, etc.

[0071] Experimental steps:

[0072] 1) Coating: Add appropriate concentration of antibody to each well of the ELISA plate and coat overnight at 4°C or room temperature.

[0073] 2) Blocking: Remove the coating solution, wash and block the plate wells with PBS or PBS containing 1-5% bovine serum albumin (BSA) or skim milk powder to reduce nonspecific binding.

[0074] 3) Sample addition: Add antigen solutions of different concentrations to each well, including negative and positive controls.

[0075] 4) Incubation: Incubate the plate at room temperature for 1 hour to allow the antigen to bind to the antibody.

[0076] 5) Washing: Wash the wells with PBS or PBS containing 0.05% Tween 20 (PBST) to remove unbound substances.

[0077] 6) Add enzyme-labeled secondary antibody: Add enzyme-labeled secondary antibody to each well and incubate for 1 hour.

[0078] 7) Wash again: Wash the plate wells to remove unbound secondary antibody.

[0079] 8) Add substrate: Add substrate solution to each well to produce color changes according to enzyme activity.

[0080] 9) Color development: Allow the color to develop for a certain period of time at room temperature in the dark.

[0081] 10) Termination reaction: Add stop solution to stop the substrate reaction.

[0082] 11) Plate reading: Use a microplate reader to read the absorbance (OD) value at an appropriate wavelength (450 nm).

[0083] 12) Curve drawing: Use GraphPad Prism software to draw a dose-response curve by plotting the antigen concentration and the corresponding average OD value.

[0084] 13) Fitting model: A four-parameter logistic regression (4PL) model was used to fit the data.

[0085] 14) Calculate EC50: Extract the EC50 value from the fitted curve, i.e., the antigen concentration that achieves 50% of the maximum response.

[0086] Example 1. Process for humanizing CD28 nanobody

[0087] This example humanizes the T cell activating factor-CD28 nanobody (VHH) by increasing the similarity between the heterologous antibody and the human antibody sequence through CDRgrafting, reducing the immunogenicity of the CD28 nanobody and maintaining its affinity for the APC antigen.

[0088] 1. Antibody framework region selection

[0089] The framework selection in this example was based on a search of the human germline and Therapeutic Structural Antibody Database (Thera-SAbDab) databases of heterologous antibodies using NCBI's blastp. The template for the humanized antibody framework region was selected based on the sequence similarity score (TOP5) of blastp (https: / / ftp.ncbi.nlm.nih.gov / blast / executables / blast+ / LATEST / ).

[0090] By using blastp to search the human germline for the amino acid sequence of the CD28 nanobody (sequence 1 in the sequence list), the sequences corresponding to the most similar germlines are shown in sequence 2-sequence 6 in the sequence list. Among them, sequence 2-sequence 5 are the germlines of the selected antibody variable region (V region), and sequence 6 is the germline of the selected antibody connecting region (J region). Finally, the sequence with the highest blastp score is selected as the human germline sequence of the CD28 nanobody for CDR region transplantation (grafting). The blastp score is the result of summing the scores of each pair of amino acid residues using a scoring matrix to score the similarity of the matching fragments. Generally speaking, the longer the matching fragment and the higher the similarity, the greater the score and the more reliable the result. The scoring results are shown in (Table 1). The thera_heavy_v_52 sequence shown in sequence 2 in the sequence list has the highest score and is used as the V region with the highest score. The highest-scoring thera_heavy_v_52 (sequence 2) and the highest-scoring J region thera_heavy_j_30 (sequence 6) were spliced ​​to obtain sequence 7, which was selected for CDR region transplantation.

[0091] Sequence 1:

[0092] QVQLVESGGGLVQAGGSLRLSCTAS ERTFRTYTMA WFRQAPGKEREFVG YITWYGGSTYYGDSVKG RFTISRDNSENTVYLEMNRLESEDTAVYYCAA GGLVSRDPGDFSS WGQGTQVTVSS.

[0093] In Sequence 1, the underlined sequence is the CDR region sequence of the CD28 nanobody, wherein the CDR1 region corresponds to amino acids 26-35 of Sequence 1, the CDR2 region corresponds to amino acids 50-66 of Sequence 1, and the CDR3 region corresponds to amino acids 99-111 of Sequence 1.

[0094] Sequence 2 (thera_heavy_v_52 (Envafolimab)):

[0095] QVQLVESGGGLVQPGGSLRLSCAASGKMSSRRCMAWFRQAPGKERERVAKLLTTSGSTYLADSVKGRFTISRDNSKNTVYLQMNSLRAEDTAVYYCAA;

[0096] Sequence 3 (IGHV3-23*03):

[0097] EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSVIYSGGSSTYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAK;

[0098] Sequence 4 (IGHV3-23*04):

[0099] EVQLVESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSAISGSGGSTYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAK;

[0100] Sequence 5 (IGHV3-64*04):

[0101] QVQLVESGGGLVQPGGSLRLSCSASGFTFSSYAMHWVRQAPGKGLEYVSAISSNGGSTYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAR;

[0102] Sequence 6 (thera_heavy_j_30 (Caplacizumab)):

[0103] GRVRTLPSEYTFWGQGTQVTVSS.

[0104] Sequence 7 (thera_heavy_v_52+thera_heavy_j_30):

[0105] QVQLVESGGGLVQPGGSLRLSCAASGKMSSRRCMAWFRQAPGKERERVAKLLLTTSGSTYLADSVKGRFTISRDNSKNTVYLQMNSLRAEDTAVYYCAAGRVRTLPSEYTFWGQGTQVTVSS.

[0106] Table 1. Blastp scoring results for the germline most similar to the CD28 antibody sequence

[0107]

[0108] 2. Selection of Sequence CDR Annotations and CDR Region Transplantation

[0109] This example uses a combined Kabat and Chothia numbering method to annotate the CDR regions in the CD28 nanobody (sequence 1) and the human germline antibody (sequence 7), respectively, to obtain the CDR regions (including CDR1-CDR3) of the CD28 nanobody (VHH). The CDR regions of the CD28 nanobody (VHH) are then replaced with the human germline: the CDR1-CDR3 regions of the CD28 nanobody (VHH) are replaced with the CDR1-CDR3 regions (including CDR1-CDR3) of the human germline antibody (sequence 7), respectively, using the abnumber package (https: / / abnumber.readthedocs.io / en / latest / ).

[0110] 2.1 Antibody sequence CDR annotation

[0111] The CDR annotation information of the human germline shown in sequence 7 was obtained using the combined numbering method of Kabat and Chothia, including the positions of the complementary determining region 1 (CDR1) sequence region of the heavy chain variable region, the antibody framework region 2 (FR2) sequence region, the complementary determining region 2 (CDR2) sequence region of the heavy chain variable region, the antibody framework region 3 (FR3) sequence region, the complementary determining region 3 (CDR3) sequence region of the heavy chain variable region, and the antibody framework region 4 (FR4) sequence region.

[0112] 2.2 CDR Region Transplantation

[0113] The sequence of the CDR1-CDR3 region of the CD28 nanobody shown in sequence 1 (the CDR1 region corresponds to amino acids 26-35 of sequence 1, the CDR2 region corresponds to amino acids 50-66 of sequence 1, and the CDR3 region corresponds to amino acids 99-111 of sequence 1) was transplanted to the CDR1-CDR3 region of sequence 2 obtained by CDR annotation in step 2.1 to obtain the CDR region transplanted sequence (sequence 8 in the sequence table). The CDR region transplanted sequence 8 was scored for similarity with sequence 1 using blastp. The results showed that the similarity parameters of the two sequences were as follows: identities = 113 / 122 (93%) (identities = 100% represents exactly the same), positives = 116 / 122 (95%) (representing similar amino acid attributes (blastp internal matrix score)), gaps = 0 / 122 (0%) (representing the difference in spaces after align).

[0114] Sequence 8:

[0115] QVQLVESGGGLVQPGGSLRLSCAASERTFRTYTMAWFRQAPGKERERVAYITWYGGSTYYGDSVKGRFTISRDNSKNTVYLQMNSLRAEDTAVYYCAAGGLVSRDPGDFSSWGQGTQVTVSS.

[0116] 3. Selection of reverse mutation sites

[0117] The affinity of antibodies for antigens may occur in more than just the CDR region. Therefore, for humanization of antibodies where the antigen-antibody binding region is located in the framework region (FR region), it is necessary to identify the affinity-related amino acid sites in the specific FR region. The present invention uses a combined scoring of biophysical and biochemical properties to easily find suitable sites.

[0118] In this example, single-point back mutations were performed on different points in the FR region of the CDR region grafted sequence based on the nanoantibody parent (alpaca) and humanized grafting (nine different amino acid points in this example, see Table 2). The open source software colabfold was then used for modeling. Then, a custom physicochemical function (based on the definition of the V region and the size of the exposed area, the consistency of the CDR region and the parent after back mutation was scored) was used to score the structural differences of the 3D structures of the CD28 nanoantibody and the 9 mutated CDR grafted protein sequences obtained by 9 different single-point back mutations.

[0119] 3.1 Sequence-to-Structure Modeling

[0120] Using the existing sequence-to-structure tool ColabFold (https: / / github.com / sokrypton / ColabFold), the nine mutant CDR-grafted protein sequences were converted to obtain the (3D) structures of the mutant CDR-grafted proteins. ColabFold was also used to obtain the 3D structure of the CD28 nanobody.

[0121] 3.2 Selection of reverse mutation sites

[0122] Humanized sequences often experience decreased affinity after transplantation alone, requiring back mutation to maintain affinity consistent with the parental sequence.

[0123] The structural difference scores of the three-dimensional structure of the CD28 nanobody and the three-dimensional structures of the nine mutated CDR-grafted protein sequences obtained by nine different single-point back mutations were scored using the following formula 1 to obtain a structural difference score:

[0124] score=important_score×exposed_area×3D_smiliarity Formula 1;

[0125] In formula 1, important_score is the numerical value obtained by scoring the similarity of the three-dimensional structure of the CD28 nanobody and the three-dimensional structure of the CDR-grafted protein sequence after mutation based on the B-factor and hydrophilicity of each amino acid residue in the sequence, which is calculated using the abnumber program package (https: / / abnumber, readthedocs.io / en / latest / ); exposed_area represents the difference in the size of the exposed area between the three-dimensional structure of the CD28 nanobody and the three-dimensional structure of the CDR-grafted protein sequence after mutation, which is calculated using the colabfold program package (https: / / github.com / sokrypton / ColabFold); 3D_similarity is the structural similarity score of the three-dimensional structure of the CD28 nanobody and the three-dimensional structure of the CDR-grafted protein sequence after mutation (representing the similarity of the protein three-dimensional structure before and after the single amino acid site back mutation), which is calculated using the pymol program package (https: / / www.pymo1.org / ).

[0126] Table 2. Structural difference scores between CD28 nanobody and humanized grafted antibodies after reverse mutation

[0127] sc reverse mutation site Score structural difference score hu_1(P14A) 1.245 hu_2(A23T) 1.267 hu_3(R47F) 0.843 hu_4(A49G) 1.276 hu_5(K76E) 1.3068 hu_6(Q82E) 1.229 hu_7(S85R) 1.3146 hu_8(R87E) 1.27 hu_9(A88S) 1.28

[0128] Note: The mutation information in the name of the back mutation site is as follows: For the back mutation site hu_1(P13A), the P in P13A means that the 13th amino acid residue in the humanized sequence mutates from P to A (i.e., the 13th amino acid residue in sequence 8 mutates from P to A).

[0129] Through calculation, the 3D structure of the CDR-grafted protein sequence after the hu_3 (R46F) site reversion mutation was obtained, and the difference score (score) between the 3D structure and the CD28 antibody structure was less than 0.85. The secondary structure difference was only two different framework sheets, and the similarity was high. Therefore, the CDR-grafted protein sequence (sequence 9) after the hu_3 (R47F) site reversion mutation of the CDR-grafted protein sequence shown in sequence 8 (i.e., the 47th amino acid residue of sequence 8 was mutated from R to F) was prepared to detect the affinity of the obtained humanized antibody to the antigen.

[0130] Sequence 9:

[0131] QVQLVESGGGLVQPGGSLRLSCAASERTFRTYTMAWFRQAPGKEREFVAYITWYGGSTYYGDSVKGRFTISRDNSKNTVYLQMNSLRAEDTAVYYCAAGGLVSRDPGDFSSWGQGTQVTVSS.

[0132] 4. Evaluation of Humanized Antibodies

[0133] The affinity between CD28 nanobody (sequence 1) and the CDR-grafted protein sequence thera_52 (sequence 9) after the hu_3 (R46F) site back mutation and CD28 was evaluated respectively.

[0134] like Figure 2 The EC50 of the parent CD28 nanobody and CD28 was 0.31 ( Figure 2 CD28-base_1 represents the CD28-base_1 in the sample); and the EC50 of thera_52 and CD28 is 0.277 ( Figure 2 CD28-base_1_hu_1 represents the middle), which is in line with expectations.

[0135] Example 2. Process for humanizing PD-L1 VHH

[0136] In this example, the nanobody against PD-L1 was humanized, and the similarity between the heterologous antibody and the human antibody sequence was increased by CDRgrafting, thereby reducing the immunogenicity of PD-L1 VHH and maintaining its affinity for the APC antigen.

[0137] 1. Antibody framework region selection

[0138] The framework selection of the present invention is based on NCBI's blastp search of the human germline of heterologous antibodies and the Therapeutic Structural Antibody Database (Thera-SAbDab), and the template of the humanized antibody framework region is selected based on the blastp score (https: / / ftp.ncbi.nlm.nih.gov / blast / executables / blast+ / LATEST / ).

[0139] Using the same process as in Example 1, the amino acid sequence of the PD-L1 Nanobody (sequence 10 in the sequence listing) was searched for the human germline using blastp, and the sequence similarity score was obtained. The sequence corresponding to the most similar germline, such as sequence 11 in the sequence listing, was used as the highest-scoring V region. The highest-scoring IGHV3-48*03 (sequence 11) and the highest-scoring J region were spliced ​​to obtain a spliced ​​sequence, and this spliced ​​sequence was selected as the human germline for CDR region transplantation to obtain the CDR region-transplanted sequence (sequence 12 in the sequence listing).

[0140] Sequence 10:

[0141] QVQLVESGGGLVQPGGSLRLSCAAS GSIFSSGTQFSDSKID WYRQAPGKQRDWIA GIFSTGSTIYEDS VKG RFAISRDNAKNMGYLQMNSLKPEDTAVYYC RVIGRGILA WGQGTQVTVSS.

[0142] In sequence 10, the underlined sequence is the sequence of the variable region CDR1-CDR3 of the PD-L1 nanobody.

[0143] Sequence 11:

[0144] IGHV3-48*03

[0145] EVQLVESGGGLVQPGGSLRLSCAASGFTFSSYEMNWVRQAPGKGLEWVSYISSSGSTIYYADSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCAR;

[0146] Sequence 12:

[0147] (PD-L1_grafted_vhh)

[0148] EVQLVESGGGLVQPGGSLRLSCAASGSIFSSGTQFSDSKIDWVRQAPGKGLEWVSGIFSTGSTIYEDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCRVIGRGILAWGQGTQVTVSS;

[0149] 2. Selection of Sequence CDR Annotations and CDR Region Transplantation

[0150] The same method as in Example 1 was used to perform CDR annotation and CDR region transplantation of the human germline sequence to obtain the CDR region transplanted sequence (sequence 12 in the sequence list). Blastp was used to score the similarity between the CDR region transplanted sequence (sequence 12 in the sequence list) and the parent PD-L1 nanobody sequence (sequence 10). The results showed that the similarity parameters of the two sequences were as follows: identities = 109 / 121 (90%) (identities = 100% represents exact identity), positives = 114 / 121 (94%) (representing similar amino acid properties (blastp internal matrix score)), gaps = 0 / 121 (0%) (representing the difference in spaces after align).

[0151] 3. Selection of reverse mutation sites

[0152] The same method as in step 3 of Example 1 was used to select the back mutation sites, specifically by aligning sequence 12 and sequence 10 to obtain 12 different amino acid sites between the CDR-grafted protein sequence and the parent PD-L1 nanobody sequence. Colabfold was used to perform structural prediction on the 12 mutated CDR-grafted protein sequences obtained by performing 12 different single-point back mutations on the CDR-grafted protein sequence (sequence 12), and the 3D structures of the 12 mutated CDR-grafted proteins were obtained.

[0153] Then, using the custom physicochemical function in Formula 1 in the example (based on the definition of the V region and the size of the exposed area, the consistency of the CDR region and the parent after back-mutation is scored), the 3D structures of the 12 mutated CDR-grafted proteins and the 3D structure of the PD-L1 nanobody sequence are scored for structural differences to obtain a structural difference score (score).

[0154] The results in Table 3 show that the five mutation sites of the CDR region grafted antibody shown in sequence 12 were reverse mutated (V42Y, G49Q, L50R, T73A and T82M) to obtain the five mutated CDR grafted proteins. The structural difference scores (scores) compared with the 3D structure of the PD-L1 nanobody sequence were small (less than 1), which were 0.967, 0.843, 0.976, 0.91 and 0.92 respectively.

[0155] Table 3. Structural difference scores between the humanized grafted PD-L1 nanobody and the PD-L1 nanobody after reverse mutation

[0156] Scanning mutation sites Score hu_1(E1Q) 1.045 hu_2(V43Y) 0.967 hu_3(G50Q) 0.843 hu_4(L51R) 0.976 hu_5(E52D) 1.3068 hu_6(V54I) 1.229 hu_7(S55A) 1.3146 hu_8(T74A) 0.91 hu_10(T83M) 0.92 hu_11(L84G) 1.27 hu_12(R92K) 1.27 hu_13(A93P) 1.28

[0157] The combination antibody hu_PD-L1 (sequence 13) with five back mutation sites (V43Y, G50Q, L51R, T74A, T83M) with a score within 1 was selected, and the same method as in Example 1 was used to prepare the antibody hu_PD-L1 with the CDR-grafted protein sequence after the corresponding five site back mutations, and the affinity of hu_PD-L1 and the PD-L1 nanobody mother to PD-L1 was tested respectively.

[0158] Sequence 13:

[0159] EVQLVESGGGLVQPGGSLRLSCAASGSIFSSGTQFSDSKIDWYRQAPGKQREWVSGIFSTGSTIYEDSVKGRFAISRDNSKNMLYLQMNSLRAEDTAVYYCRVIGRGILAWGQGTQVTVSS(hu_PD-L1).

[0160] The results are as follows Figure 3 As shown in the SPR (surface plasmon resonance) full kinetic analysis, the affinity determination results of the PD-L1 nanobody parent showed that its binding constant (Kd) with the PD-L1 protein was 4.04E-9 (as shown in Figure 3 In contrast, the SPR test results of the hu_PD-L1 nanobody showed that its binding constant (Kd) to PD-L1 was 1.44E-8 (as shown in Figure 3 Based on this data, we can observe that the affinity of the hu_PD-L1 nanobody to PD-L1 is slightly lower than that of the parent nanobody, while retaining the high affinity of the parent antibody, which is consistent with the expectations of a humanized antibody.

[0161] The present invention has been described in detail above. For those skilled in the art, without departing from the purpose and scope of the present invention, and without the need to carry out unnecessary experimental conditions, the present invention can be implemented in a wide range under equivalent parameters, concentrations and conditions. Although the present invention provides specific embodiments, it should be understood that further improvements can be made to the present invention. In short, according to the principles of the present invention, this application is intended to include any changes, uses or improvements to the present invention, including changes that depart from the disclosed scope in this application and are made using conventional techniques known in the art.

Claims

1. A device obtained based on a humanized antibody with CDR region grafting, characterized in that: The device includes the following modules: A1) Humanized framework region selection module: used to obtain humanized antibody sequences of heterologous antibodies specific for the target antigen using sequence alignment software; A2) CDR region transplantation module: used to annotate the antigen complementary determining region sequence of the heterologous antibody using an antibody sequence CDR annotation system to obtain a heterologous antibody CDR sequence, and to annotate the humanized antibody sequence to obtain a humanized antibody CDR sequence; and to replace the humanized antibody CDR sequence in the humanized antibody sequence with the heterologous antibody CDR sequence to obtain a CDR region transplanted antibody sequence for the target antigen; A3) A module for selecting back mutation sites and obtaining humanized antibodies: used to align the sequence of the CDR region-grafted antibody with the sequence of the heterologous antibody to obtain differential amino acid sites, perform back mutation on each of the differential amino acid sites to obtain single amino acid back mutation CDR region-grafted antibodies, perform structural prediction on all of the single amino acid back mutation CDR region-grafted antibodies using protein structure prediction software to obtain the three-dimensional structures of all of the single amino acid back mutation CDR region-grafted antibodies, and use the protein structure prediction software to predict the three-dimensional structure of the heterologous antibody to obtain the three-dimensional structure of the heterologous antibody; The three-dimensional structure of the single amino acid back-mutated CDR region transplanted antibody and the three-dimensional structure of the heterologous antibody are scored for structural difference using the following formula 1 to obtain a structural difference score; Formula 1; In Formula 1, score represents the structural difference score between the three-dimensional structure of the single amino acid back-mutated CDR region-grafted antibody and the three-dimensional structure of the heterologous antibody; important_score is the value obtained by scoring the similarity between the three-dimensional structure of the single amino acid reversion mutation CDR region grafted antibody and the three-dimensional structure of the heterologous antibody based on the B-factor and hydrophilicity of each amino acid residue in the sequence; exposed_area represents the difference in the size of the exposed area of ​​the three-dimensional structure of the heterologous antibody and the three-dimensional structure of the single amino acid reversion mutation CDR region grafted antibody; 3D_similarity is the structural similarity score between the three-dimensional structure of the single amino acid reversion mutation CDR region grafted antibody and the three-dimensional structure of the heterologous antibody; The amino acid mutation sites in the single amino acid back mutation CDR region transplanted antibody with a structural difference score less than 1 are selected and merged to obtain back mutation combination sites, and the amino acid residues at the back mutation combination site position in the CDR region transplanted antibody sequence are replaced with the amino acid residues at the back mutation combination site position corresponding to the heterologous antibody to obtain a humanized antibody sequence for the target antigen.

2. The device according to claim 1, characterized in that: The heterologous antibody is a non-human antibody.

3. The device according to claim 1 or 2, characterized in that: A3) In formula 1, the important_score is calculated using the abnumber program package; the exposed_area is calculated using the colabfold program package; and the 3D_similarity is calculated using the pymol program package.

4. The device according to claim 1 or 2, characterized in that: The antibody sequence CDR annotation system is a CDR annotation system that combines Kabat and Chothia numbering.

5. A method for obtaining a humanized antibody based on CDR region transplantation, characterized in that: The method comprises the following steps: B1) Selection of humanized framework regions: Use sequence alignment software to obtain the humanized antibody sequence of the target antigen heterologous antibody; B2) CDR region transplantation: using an antibody sequence CDR annotation system to annotate the antigen complementary determining region sequence of the heterologous antibody to obtain a heterologous antibody CDR sequence, and annotating the humanized antibody sequence to obtain a humanized antibody CDR sequence; replacing the humanized antibody CDR sequence in the humanized antibody sequence with the heterologous antibody CDR sequence to obtain a CDR region transplanted antibody sequence for the target antigen; B3) Selection of back mutation sites and obtaining humanized antibodies: comparing the sequence of the CDR region-grafted antibody with the sequence of the heterologous antibody to obtain differential amino acid sites, performing back mutation on each of the differential amino acid sites to obtain single amino acid back mutation CDR region-grafted antibodies, performing structural prediction on all of the single amino acid back mutation CDR region-grafted antibodies using protein structure prediction software to obtain the three-dimensional structures of all of the single amino acid back mutation CDR region-grafted antibodies, and using the protein structure prediction software to predict the three-dimensional structure of the heterologous antibody to obtain the three-dimensional structure of the heterologous antibody; The three-dimensional structure of the single amino acid back-mutated CDR region transplanted antibody and the three-dimensional structure of the heterologous antibody are scored for structural difference using the following formula 1 to obtain a structural difference score; Formula 1; In Formula 1, score represents the structural difference score between the three-dimensional structure of the single amino acid back-mutated CDR region-grafted antibody and the three-dimensional structure of the heterologous antibody; important_score is the value obtained by scoring the similarity between the three-dimensional structure of the single amino acid reversion mutation CDR region grafted antibody and the three-dimensional structure of the heterologous antibody based on the B-factor and hydrophilicity of each amino acid residue in the sequence; exposed_area represents the difference in the size of the exposed area of ​​the three-dimensional structure of the heterologous antibody and the three-dimensional structure of the single amino acid reversion mutation CDR region grafted antibody; 3D_similarity is the structural similarity score between the three-dimensional structure of the single amino acid reversion mutation CDR region grafted antibody and the three-dimensional structure of the heterologous antibody; The amino acid mutation sites in the single amino acid back mutation CDR region transplanted antibody with a structural difference score less than 1 are selected and merged to obtain back mutation combination sites, and the amino acid residues at the back mutation combination site position in the sequence of the CDR region transplanted antibody are replaced with the amino acid residues at the back mutation combination site position corresponding to the heterologous antibody to obtain a humanized antibody sequence for the target antigen.

6. The method according to claim 5, characterized in that: The heterologous antibody is a non-human antibody.

7. The method according to claim 5 or 6, characterized in that: The antibody sequence CDR annotation system is a CDR annotation system that combines Kabat and Chothia numbering.

8. The method according to claim 5 or 6, characterized in that: In formula 1 of B3), the important_score is calculated using the abnumber program package; in formula 1 of B3), the exposed_area is calculated using the colabfold program package; in formula 1 of B3), the 3D_similarity is calculated using the pymol program package.

9. A computer-readable storage medium storing a computer program, characterized in that: The computer program enables a computer to execute the steps of the method according to any one of claims 5 to 8.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory, wherein: The processor executes the computer program to implement the steps of the method according to any one of claims 5 to 8.

Images