Bispecific antibody targeting heterogeneous macrophages

Abstract

The invention discloses a bispecific antibody targeting heterogeneous macrophages, and relates to the field of biological medicines. According to the present invention, bevacizumab is adopted as a parent antibody skeleton, antibody arms targeting CD16 are arranged on the N end side, antibody arms targeting CD163 are located on the C end side, knob mutation (K322A and T366W) is introduced into the heavy chain CH3 structural domain targeting CD16, three complementary hole mutations (T366S, L368A and Y407V) are introduced into the heavy chain CH3 structural domain targeting CD163, and cysteine residues are inserted into the S354 site of the knob chain and the Y349 site of the hole chain; variable regions (V regions) targeting CD16 and CD163 are connected in series through a (G4S) 3 flexible linker and then fused with respective constant regions (C regions). The double-antibody architecture disclosed by the invention realizes collaborative optimization in structural stability, double-antigen binding specificity, macrophage targeting and functional activity, provides an efficient and safe novel candidate drug for targeted therapy of related diseases, and has important clinical transformation value and application prospect.

Description

Technical Field

[0001] This invention relates to the field of biomedicine, specifically a bispecific antibody that targets heterogeneous macrophages. Background Technology

[0002] Macrophages are a type of innate immune cell originating from hematopoietic stem cells in the bone marrow and widely distributed in various tissues and organs of the body. They have three main functions: phagocytosis, presentation of exogenous antigens, and immune regulation. Their significant heterogeneity is the core basis for their diverse biological functions. This heterogeneity is mainly reflected in the plasticity of their phenotype, function, and regulatory mechanisms. The core characteristic is the polarization phenomenon that occurs after stimulation by microenvironmental signals. They are mainly divided into two types: pro-inflammatory M1 macrophages and anti-inflammatory M2 macrophages.

[0003] The various activation stages of macrophages are crucial for the inflammatory evolution of atherosclerosis, metabolic homeostasis, allergic diseases, autoimmune diseases, type 2 diabetes, and other inflammatory diseases. During inflammation, two types of macrophages are activated alternately. M1-like macrophages dominate in the early stages, sending pro-inflammatory signals such as interleukins to participate in pathogen clearance, limit cell proliferation, and cause tissue damage. Subsequently, M2-like macrophages secrete more and release repair mediators, which help promote tissue remodeling and anti-inflammatory responses. When macrophage polarization is disrupted, it leads to an imbalance in the M1 / M2 ratio, with excessive M1 activation accompanied by insufficient M2 repair function, thereby exacerbating inflammation. In the pathological process of neurodegenerative diseases (such as Alzheimer's disease), the imbalance in the polarization state of microglia in the brain, as macrophage-like innate immune cells, affects disease progression. Activated M1 microglia secrete large amounts of pro-inflammatory cytokines, exacerbating the abnormal aggregation and deposition of β-amyloid protein and directly inducing oxidative stress damage and apoptosis in neurons. Conversely, M2 microglia possess highly efficient phagocytic clearance capabilities for abnormal proteins and can mediate the repair and functional remodeling of damaged neural tissue through the secretion of neurotrophic factors. Meanwhile, tumor-associated macrophages, as a core component of the tumor microenvironment, directly regulate tumor progression due to their significant polarization heterogeneity. In the tumor microenvironment, M1 macrophages exert anti-tumor effects, mediating phagocytosis and antibody-dependent cell-mediated cytotoxicity to kill tumor cells. Conversely, M2 macrophages can inhibit T cell-mediated anti-tumor immune responses, promote tumor angiogenesis, and accelerate tumor progression and metastasis.

[0004] In summary, the heterogeneity of macrophages underscores their crucial role in disease pathogenesis, making them appropriate targets for alleviating or treating various inflammatory diseases. Therefore, precise targeting of heterogeneous macrophages will play a significant role in alleviating or curing a variety of diseases.

[0005] Currently, methods for targeting macrophages include small molecule regulators, nanodelivery systems, and monoclonal antibody-mediated targeted therapies.

[0006] Small molecule regulators are therapeutic methods that target macrophage surface receptors or intracellular signaling pathways to achieve functional regulation or drug delivery. They possess core advantages such as strong penetration, ease of mass production, and controllable pharmacokinetics. Mannose is a natural ligand for the macrophage mannose receptor (CD206), and the binding between the two exhibits high molecular specificity. By covalently coupling mannose with the chemotherapeutic drug doxorubicin via a pH-sensitive bond, a mannose-doxorubicin small molecule conjugate is constructed. The conjugate undergoes specific receptor-ligand binding with CD206 on the macrophage surface, forming a complex that triggers receptor-mediated endocytosis on macrophages. This complex encapsulates the conjugate into endosome vesicles, which are then transported to lysosomes. The acidic environment of the lysosome (pH 4.5-5.5) breaks the pH-sensitive bond of the conjugate, releasing free doxorubicin and achieving precise intracellular drug delivery to macrophages. However, this method has limitations in target specificity; the targeting range is restricted by receptor expression, and it can only target macrophage subsets that highly express CD206.

[0007] Nanoparticle delivery systems utilize the passive targeting (EPR effect) or active targeting (ligand modification) properties of nanoscale carriers to precisely deliver therapeutic drugs to macrophages. Their core advantages include a wide drug loading range, good biocompatibility, and reduced systemic drug toxicity. However, this method suffers from problems such as easy inactivation of targeting ligands, broad specificity, and insufficient targeting depth.

[0008] Monoclonal antibody-mediated targeted therapy utilizes the highly specific binding ability of monoclonal antibodies to precisely regulate the function, location, or state of macrophages. It can be mainly divided into two core pathways of action: one is to directly bind to function-related receptors on the surface of macrophages, such as CCR2 and SIRPα, and regulate the differentiation, migration, phagocytosis, or polarization phenotype of macrophages by blocking or activating the corresponding signaling pathways; the other is to use monoclonal antibodies as targeting carriers, conjugate them with therapeutic substances, and bind to specific markers on the surface of macrophages (such as CD68 and CD11b). Through the endocytosis of macrophages, the therapeutic substances are precisely delivered into the interior of macrophages to achieve the clearance of abnormal macrophages or the regulation of their function.

[0009] Regarding the few macrophage-selective chemotherapeutic agents mentioned above, their limitations in disease intervention are mainly as follows: First, most of these agents can only act on a single pathway or dimension regulated by macrophages, failing to meet the needs of multi-pathway synergistic regulation by macrophages in complex pathological microenvironments. Second, these agents lack precise cell selectivity, interfering with other normal cells while intervening in macrophages. For example, SGLT2 inhibitors like dapagliflozin, while regulating macrophage glycolysis by inhibiting PFKFB3 enzyme activity, also affect the glucose metabolism of other normal cells. Furthermore, due to insufficient tissue targeting, they cannot precisely accumulate around macrophages at pathological sites. After systemic administration, they are widely distributed throughout the body, weakening the drug concentration at the pathological site and reducing therapeutic efficacy, while also damaging normal macrophages and weakening the body's innate immune capacity. Furthermore, some macrophage-selective chemotherapeutic agents that act on metabolic pathways have a short duration of action, requiring frequent administration to maintain their regulatory effect on macrophages. Frequent administration can easily lead to drug accumulation in the body, increasing the risk of adverse reactions. For example, some chemotherapeutic agents that regulate glycolysis may interfere with the body's energy metabolism homeostasis and cause problems such as abnormal blood sugar levels if they accumulate over a long period of time.

[0010] While monoclonal antibodies can regulate macrophage function by specifically binding to certain markers on the surface of macrophages, playing a unique role in inflammatory diseases and tumor microenvironment intervention, this class of drugs still has many limitations: First, the therapeutic limitation of single target; these drugs can only bind to one antigenic epitope on the surface of macrophages. Macrophages exhibit heterogeneity in phenotype and function. For example, macrophage monoclonal antibodies targeting CD163 may experience reduced drug binding capacity or even fail to exert their regulatory effect due to decreased CD163 expression abundance in tumor-associated macrophages. Second, the relative simplicity of function and effect; natural macrophage-targeting monoclonal antibodies mainly regulate macrophages through Fc-mediated effects, making it difficult to simultaneously regulate multiple functions of macrophages such as polarization, phagocytosis, and cytokine secretion, thus limiting their intervention effect on complex pathological processes involving macrophages.

[0011] In the field of bispecific antibody construction, existing technologies mostly adopt conventional architectures for design. Some bispecific antibodies suffer from heavy chain mismatch and light chain cross-binding disorder, resulting in insufficient antibody stability. At the same time, most bispecific antibodies retain the FcγR-mediated antibody-dependent cytotoxicity (ADCC) effect, which can easily cause damage to non-target cells and pose an off-target risk. In addition, some bispecific antibody architectures have insufficient binding affinity to the target, making it difficult to achieve efficient targeted recognition. Summary of the Invention

[0012] In view of the many shortcomings of existing treatments for macrophage-mediated diseases, this invention aims to design a CD16-CD163 bispecific antibody targeting heterogeneous macrophages and its construction method, so as to solve the problems of insufficient target precision, single dimension of action, unstable bispecific antibody structure, and insufficient targeting safety of macrophage-targeted therapy.

[0013] In view of this, the present invention designs a CD16-CD163 bispecific antibody targeting heterogeneous macrophages. By recognizing two different receptors, it achieves broad-spectrum coverage of the heterogeneous macrophage population, transforming the understanding of macrophage heterogeneity and the M1 / M2 functional subtyping mechanism into a targeted therapeutic tool.

[0014] This invention innovatively proposes a bispecific antibody construction strategy targeting CD16 and CD163. The core of the invention is a bispecific antibody-targeted macrophage therapy, relying on the dual-target binding characteristics of the two antibodies to precisely target M1 and M2 macrophages. CD16 is the FcγRⅢ receptor on the macrophage surface, an activating receptor that, upon binding to its ligand, activates pro-inflammatory and phagocytic signaling in macrophages, making it a key target for inducing macrophage polarization towards M1. CD163 is a scavenger receptor on the macrophage surface, highly expressed only on anti-inflammatory M2 macrophages, and also a regulatory molecule for macrophage anti-inflammatory function. The synergistic effect of these two antibodies efficiently transforms M2 macrophages, which are originally in an anti-inflammatory state, into pro-inflammatory M1 macrophages with phagocytic and killing capabilities.

[0015] The technical solution adopted in this invention is:

[0016] A bispecific antibody targeting heterogeneous macrophages uses bevacizumab as the parental antibody backbone. The antibody arm targeting CD16 is located on the N-terminal side, and the antibody arm targeting CD163 is located on the C-terminal side. A "knob" mutation (K322A, T366W) is introduced into the CH3 domain of the heavy chain targeting CD16, while three complementary "hole" mutations (T366S, L368A, Y407V) are introduced into the CH3 domain of the heavy chain targeting CD163. Cysteine ​​residues are inserted at position S354 of the "knob chain" and position Y349 of the "hole chain". The variable regions (V regions) targeting CD16 and CD163 are linked in series through a (G4S)3 flexible linker and then fused with their respective constant regions (C regions).

[0017] Specifically, LALA mutations (L254A, L255A) are introduced into the CH2 domain of the Fc region.

[0018] This invention also provides a method for designing bispecific antibodies targeting heterogeneous macrophages, comprising the following steps:

[0019] Screening for M1 / M2 type AMs-specific surface markers: Differential gene expression analysis was performed on single-cell sequencing datasets of AMs from public databases to screen CD16 and CD163 as the most important targets;

[0020] In the bispecific antibody architecture constructed using the Kih / CrossMab-LALA technology, bevacizumab (IgG1κ type) was selected as the parental antibody backbone. The antibody arms targeting CD16 were all located on the N-terminal side, and the antibody arms targeting CD163 were located on the C-terminal side. The "Knobs-into-Holes" technology was used to exchange the heavy chain CH1 domain and the light chain CL domain targeting CD163.

[0021] A dual-resistance architecture is constructed using DVD-Ig technology: the variable regions (V regions) of the target CD16 and CD163 are connected in series through (G4S)3 flexible connectors, and then fused with their respective constant regions (C regions).

[0022] Furthermore, the specific method of exchanging the heavy chain CH1 domain and light chain CL domain targeting CD163 involves introducing a "knob" mutation (K322A, T366W) into the heavy chain CH3 domain targeting CD16, while simultaneously introducing three complementary "hole" mutations (T366S, L368A, Y407V) into the heavy chain CH3 domain targeting CD163.

[0023] Furthermore, cysteine ​​residues are inserted at position S354 of the “knob chain” and position Y349 of the “hole chain”.

[0024] Furthermore, the sequence of the (G4S)3 flexible connector is: GGGGSGGGGSGGGGS.

[0025] Finally, this invention provides a method for preparing a bispecific antibody targeting heterogeneous macrophages, comprising the following steps:

[0026] Transformation: Take 50 μL of competent E. coli, add 2 μg of target plasmid (corresponding to the four strands of vector respectively), and incubate on ice for 30 minutes; the target plasmid includes at least one set of nucleotide sequence fragments: SEQ ID NO.1 and SEQ ID NO.2, SEQ ID NO.5 and SEQ ID NO.6, SEQ ID NO.9 and SEQ ID NO.10, SEQ ID NO.13 and SEQ ID NO.14.

[0027] Resuscitation: Add 500 μL of preheated LB medium to the transformed bacterial culture and incubate at 37°C in a shaker for 1 hour.

[0028] Screening and sequencing: 100 μL of revived bacterial culture was spread on LB agar plates containing ampicillin and incubated upside down at 37°C overnight. The next day, single colonies were picked and inoculated into 1.5 mL of LB liquid medium containing antibiotics and cultured with shaking at 37°C. Plasmids were extracted and sent for sequencing. The sequences were compared to confirm that the gene sequences were correct.

[0029] Expanded culture: Inoculate the correctly sequenced bacterial culture into 25 mL LB liquid medium and incubate at 37°C in a shaker for 12-16 hours.

[0030] Plasmid extraction: Plasmids were extracted using an endotoxin-free plasmid extraction kit.

[0031] Using the Expi293 expression system, the four plasmids prepared above were transfected into HEK293F cells for expression.

[0032] The transfection complex was added dropwise to the cell suspension, and the cells were placed in a 37°C, 8% CO2 shaker incubator for suspension culture. The antibody-containing cell culture supernatant was harvested by centrifugation at 14000×g for 30 minutes, purified, and then filtered through a 0.22μm sterile filter.

[0033] The beneficial technical effects of this invention include:

[0034] 1. Existing monoclonal antibodies, chemotherapy agents and other therapies lack subtype specificity in target selection, and are prone to acting on markers co-expressed by normal macrophages and pathological macrophages, causing off-target effects and toxic side effects. This invention achieves specific targeting of diseased alveolar macrophages through precise target selection, reducing damage to normal macrophages.

[0035] 2. Most existing macrophage-targeted therapies can only act on a single pathway or function regulated by macrophages, and cannot adapt to the complex pathological mechanisms mediated by the heterogeneity of alveolar macrophage M1 and M2 subtypes. This invention achieves dual-target recognition of M1 / M2 alveolar macrophages, covering the heterogeneous population of AMs, and solves the "incomplete target" defect of traditional single-target tools.

[0036] 3. In the dual-antibody architecture constructed based on Kih / CrossMab-LALA technology, stability is improved by "Knobs-into-Holes" heavy chain pairing. At the same time, the CH1-CL region CrossMab cross-interchange of the Fab domain of CD163 solves the problems of light chain mismatch and light chain disorder, thereby improving the stability and assembly accuracy of the dual-antibody architecture.

[0037] 4. Introduce PG-LALA mutations in the Fc region to eliminate FcγR-mediated ADCC effects, reduce off-target damage risk, and improve targeting safety.

[0038] 5. By optimizing the structure, the binding affinity between the bispecific antibody and CD16 and CD163 targets is improved, achieving efficient and precise targeted binding.

[0039] 6. The bispecific antibody of this invention can form an "immune synapse-like" bridging structure between M1 / M2 type AMs. By inhibiting TNF-α signaling of M1 cells and activating the JAK-STAT pathway of M2 cells, it can achieve bidirectional dialogue and polarization homeostasis reconstruction between subpopulations. This mechanism suggests that macrophage subpopulations do not act in isolation, but form a stable immune interaction network through physical-molecular bridging.

[0040] In summary, this invention successfully constructed a bispecific antibody targeting CD16 and CD163 by combining CrossMab technology with Knob-in-Hole (KiH) mutation and interchain disulfide bond modification. Compared with traditional bispecific antibodies such as DVD-Ig and monoclonal antibodies, it has significant technical advantages and application value: Structurally, the monomer purity of this bispecific antibody was verified to be 99.2% by SEC-HPLC. Molecular docking analysis showed that its HADDOCK score was as low as -152±3, significantly better than DVD-Ig (-118±5, ***P<0.001). Key indicators such as cluster size and constraint violation energy all showed better conformational stability and energy adaptability. In terms of binding activity, ELISA experiments confirmed that its binding affinity for CD16 and CD163 reached 0.9 nM and 0.8 nM, respectively, which is comparable to the activity of corresponding single-target monoclonal antibodies, enabling specific and simultaneous binding of two antigens. Functionally, flow cytometry analysis showed that it has a specific binding affinity for M1 / M2. The binding rate of type 1 alveolar macrophages reached 47.8±3.1%, which was significantly higher than that of CD16 monoclonal antibody (33.6±1.2%) and CD163 monoclonal antibody (26.3±0.8%) (P<0.001).

[0041] Based on the confirmed structural stability, dual antigen binding specificity, and macrophage targeting of the bispecific antibody architecture of this invention, its biological function was further verified through animal experiments. Fluorescein-labeled bispecific antibodies were nasally sprayed into the airways of mice, and their dynamic distribution and organ targeting in mice were successfully detected. Results showed that the bispecific antibody could precisely target lung tissue with a targeting rate of 96.78%, and its retention time in lung tissue was as long as 120 hours. Confocal microscopy confirmed that BsAb could form an "immunosynapse-like" bridging structure between M1 and M2 type AMs. Flow cytometry results further corroborated that the binding ratio and binding amount of the CD16-CD163 bispecific antibody to macrophages were higher than those of the CD16 / CD163 monoclonal antibody. Figure 6 As shown. In summary, the dual-antibody architecture of this invention achieves synergistic optimization in structural stability, dual-antigen binding specificity, macrophage targeting, and functional activity, providing a highly efficient and safe novel drug candidate for targeted therapy of related diseases, and has significant clinical translational value and application prospects. Attached Figure Description

[0042] Figure 1 Data graph for screening specific surface markers of M1 / M2 type AMs;

[0043] In the figure, A: Differential expression profiles of characteristic genes of M1 / M2 type AMs, UMAP visualization based on GEO single-cell dataset (GSE269205): differentially expressed genes of M1 type (red) and M2 type (yellow) AMs (|log2FC|>1, Pvalue<0.05); B: Specificity analysis of M1 / M2 macrophage markers, CellMarker database: (ii) Venn diagram of common and unique genes of M1 / M2; (i) (ii) Word cloud plots showing CD16 (FCGR3A) as an M1-specific marker and CD163 as an M2-specific marker, respectively; C: Spatial expression verification of CD16 / CD163 in inflamed lung tissue, HPA immunohistochemistry: CD16 is highly expressed in the M1 polarization region (left) and CD163 (right) is highly expressed in the M2 polarization region (scale bar: 100μm).

[0044] Figure 2 Diagram showing the design and architecture construction of bispecific antibodies;

[0045] In the figure, D: Kih / CrossMab-LALA bispecific antibody structural design, tertiary (i) and secondary (ii) structural models: Kih technology mediates heavy chain (HC) pairing (Fc segment), CD163 Fab domain CH1 / CL cross-interchange (light red and light blue); E: DVD-Ig bispecific antibody structural design, tertiary (i) and secondary (ii) structural models: N-terminal CD16 VH-VL (cyan + yellow), C-terminal CD163 native Fab (green), dual variable domains tandemly linked to the IgG backbone; F: Molecular docking simulation of CrossMab bispecific antibody, visualized after Alphafold3 prediction: (i) CD16 antigen binding interface with CrossMab; (ii) CD163 antigen binding interface with CrossMab; hydrogen bonds are marked with yellow dashed lines; G: Molecular docking simulation of DVD-Ig bispecific antibody, visualized after Alphafold3 prediction: (i) CD16 binding interface with DVD-Ig; (ii) CD163 binding interface with DVD-Ig.

[0046] Figure 3 For a multidimensional evaluation of the efficacy of dual-antibody binding;

[0047] In the figure, (i) cluster size; (ii) Z-score (CrossMab: -2.3 vs DVD-Ig: -1.1); (iii) constraint violation energy (ΔE=12.7 kcal / mol); (iv) desolvation energy (CrossMab: -8.9 kcal / mol); (v) HADDOCK score (CrossMab: -152±3 vs DVD-Ig: -118±5; ***P<0.001).

[0048] Figure 4 The results are the verification results of the bispecific antibodies prepared in this invention;

[0049] In the figure, I: Bispecific antibody integrity verification (SDS-PAGE), electrophoresis: the reduced (R) Kih / CrossMab antibody shows 4 bands, corresponding to CD163 heavy chain, CD16 heavy chain, CD163 light chain, and CD16 light chain (lane 1), and the non-reducing band shows a single band (lane 2) (150±3 kDa); J: Bispecific antibody purity analysis (SEC-HPLC), chromatogram: the main peak retention time of Kih / CrossMab is 2.5±0.2 min (purity 97.8%).

[0050] Figure 5 The biological characteristics of the antibodies prepared in this invention;

[0051] In the figure, K: Bispecific antibody antigen binding affinity assay, (i) ELISA dose curve: CD16 monoclonal antibody vs CrossMab bispecific antibody (ECG) 50 =0.9nM); (ii) CD163 monoclonal antibody vs CrossMab bispecific antibody (EC) 50 =0.8 nM)(n=3); L: Establishment of macrophage polarization model, (i) M1 type induction procedure: LPS+IFN-γ stimulation (CD11b+CD68+CD86+ cells accounted for 47.2±1.33%); (ii) M2 type induction procedure: IL-4+Il13 stimulation (CD11b+CD68+CD206+ cells accounted for 55.4±1.76%); M: Flow cytometry verification of bispecific antibody targeting binding ability, (i) Flow cytometry scatter plot (smooth), (ii) Peak plot comparison: binding rate of bispecific antibody group 47.8±3.1% vs monoclonal antibody group (CD16: 33.6±1.2%; CD163: 26.3±0.8%; ***P<0.001).

[0052] Figure 6 Animal experiments were conducted to verify the preparation of antibodies for this invention.

[0053] In the figure, DE: Fluorescein-labeled bispecific antibodies were used to assess the biodistribution and in vivo retention of the bispecific antibodies after intranasal administration; A: Confocal microscopy images show the BsAb-mediated (red, labeled with Zenon) linkage phenomenon, namely M1-AMs (green, labeled with CellTrace). TM CFSE cell markers) and M2-AMs (blue, using CellTrace) TM (CTV marker). The right side shows the quantitative analysis results of intercellular distance and connection efficiency. B: Flow cytometry analysis of M1-M2 macrophage connections. Gating strategies and histograms illustrate the bridging efficiency mediated by BsAb compared to the control monoclonal antibody (mAb). Scale bar: 2 μm. Detailed Implementation

[0054] To achieve the above objectives, the present invention adopts the following technical solution. The overall process includes five core stages: specific target screening, bispecific antibody architecture design, bispecific antibody architecture AI optimization, antibody preparation and in vitro expression and purification, and biological characteristic verification. The steps are as follows:

[0055] Step 1: Screening for M1 / M2 type AMs specific surface markers using bioinformatics analysis techniques: Single-cell sequencing datasets of AMs were collected from public databases, and differential gene expression analysis was performed. The results showed that macrophages are not strictly divided into M1 and M2 subpopulations, but rather there are a large number of intermediate cells with coexisting M1 / M2 polarization characteristics. Specific analysis of M1 / M2 macrophage markers, combined with the CellMarker database, identified 7 genes shared by M1 and M2. Among these, 45 genes were specific to the M1 type AMs subpopulation, with CD16 showing the highest expression. The M2 type AMs subpopulation had 34 genes specific, with CD163 being the most prominent. Therefore, candidate surface markers meeting the criteria for high expression in M1 / M2 type AMs were initially screened. Further immunohistochemical analysis of pneumonia tissues in the HumanProtein Atlas database was conducted. By comparing the expression and localization of candidate markers in normal and inflamed tissues, it was verified that CD16 is highly expressed in the M1 polarization region, and CD163 is highly expressed in the M2 polarization region, ultimately determining the precise target. Figure 1 As shown.

[0056] Step 2: Design and construct bispecific antibodies for the specific biomarkers obtained through screening, such as... Figure 2As shown: Referring to the reported antigen-binding domain sequence design principles, based on the extracellular domain structure of the target protein, the antigen-antibody binding interface was analyzed using modeling software. Two bispecific antibody molecular architectures were constructed using Kih / CrossMab-LALA technology and Dual Variable Domain Immunoglobulin (DVD-Ig) technology, respectively. ① In the bispecific antibody architecture constructed using Kih / CrossMab-LALA technology, bevacizumab (IgG1κ type) was selected as the parental antibody backbone, simultaneously binding human anti-CD16 antibody and anti-CD163 antibody. In all constructed bispecific antibodies, the antibody arm targeting CD16 was located on the N-terminal side (left arm), while the antibody arm targeting CD163 was located on the C-terminal side (right arm). To enhance structural stability and promote heavy chain-specific heterodimerization, a "Knobs-into-Holes" technique was employed, involving domain exchange between the CH1 domain of the heavy chain targeting CD163 and the CL domain of the light chain. Specifically, "knob" mutations (K322A, T366W) were introduced into the CH3 domain of the heavy chain targeting CD16, while three complementary "hole" mutations (T366S, L368A, Y407V) were introduced into the CH3 domain of the heavy chain targeting CD163. Furthermore, cysteine ​​residues were strategically inserted at position S354 of the "knob chain" and position Y349 of the "hole chain" to form stable interchain disulfide bonds, further enhancing heavy chain pairing. To eliminate antibody-dependent cytotoxicity (ADCC), LALA mutations (L254A, L255A) were introduced into the CH2 domain of the Fc region; ② The dual-variable domain immunoglobulin (DVD-Ig) architecture constructed using dual-variable domain immunoglobulin (DVD-Ig) technology connects the variable regions (V regions) targeting CD16 and CD163 in series via a (G4S)3 flexible linker (sequence: GGGGSGGGGSGGGGS), and then fuses them with their respective constant regions (C regions). This structure can form a bispecific antibody architecture that can simultaneously and specifically bind to CD16 and CD163, achieving dual-targeting function.

[0057] Step 3: Optimize the bispecific antibody architecture using AI molecular simulation technology: Three-dimensional structural models of two bispecific antibody architectures were constructed using Alphafold3 software. The input sequences included the variable domains (VH, VL) of the antibody and the extracellular domains of CD16 / CD163. Protein-protein docking was performed using HADDOCK 2.4 to simulate the molecular docking between the antibody and the two targets. The docking was performed in two steps: first, the VH / VL domain targeting CD16 was docked with CD16, and then the VH / VL domain targeting CD163 was docked with CD163. Bidirectional interaction constraints were defined based on the interface residues predicted from the AlphaFold model. The docking conformations were clustered using the common contact fraction (FCC) criterion. The "cluster size" of CrossMab was significantly higher than that of DVD-Ig, directly demonstrating that CrossMab can form a more stable and dominant conformation during binding. Meanwhile, the HADDOCK score of CrossMab was -152±3, significantly lower than that of DVD-Ig (-118±5), with a statistically significant difference of ***P<0.001. Next, the interfacial properties and binding free energy of the antibodies were analyzed using a PDBePISA server. Calculated parameters included solvation free energy gain (ΔG), interfacial area, number of hydrogen bonds, and salt bridges to assess antibody stability and binding affinity. A comparative analysis of the CrossMab and DVD-Ig formats was conducted to determine the structural determinants of differential antigen recognition. Docking results and binding energy data were processed using an internal Python script. Hydrogen bonds, hydrophobic contacts, and electrostatic interactions at the antibody-antigen interface were visualized in PyMOL to screen for the optimal bispecific antibody architecture with a more significant targeted binding advantage. The results showed that the Z-score of CrossMab (-2.3) was lower than that of DVD-Ig (-1.1); the constraint violation energy of CrossMab was significantly lower than that of DVD-Ig; and the desolvation energy of CrossMab was -8.9 kcal / mol. In summary, the Kih / CrossMab structure exhibits superior conformational stability and energy adaptability, and the statistical significance of the HADDOCK score further demonstrates its significantly stronger binding efficiency compared to DVD-Ig. The results are as follows: Figure 3 As shown.

[0058] Step 4: Antibody preparation and in vitro expression and purification: In order to obtain high-purity, endotoxin-free transfection-grade plasmids, four expression plasmids encoding two different heavy chains and two different light chains of CrossMab were prepared according to the steps of transformation-resuscitation-screening and sequencing-expansion culture-plasmid extraction.

[0059] Transformation: Take 50 μL of competent E. coli, add 2 μg of the target plasmid (corresponding to the four strands of the vector; see the antibody sequence below for specific sequences), and mix gently. Incubate on ice for 30 minutes. Heat shock at 42°C for 1 minute, then quickly transfer to ice and incubate for 5 minutes.

[0060] Resuscitation: Add 500 μL of preheated LB medium (antibiotic-free) to the transformed bacterial culture. Incubate at 37°C with shaking at 220 rpm for 1 hour.

[0061] Screening and sequencing: 100 μL of revived bacterial culture was spread onto LB agar plates containing ampicillin and incubated upside down at 37°C overnight. The next day, single colonies were picked and inoculated into 1.5 mL of LB liquid medium containing antibiotics and cultured with shaking at 37°C. A small amount of plasmid was extracted and sent for sequencing; the sequence was compared to confirm that the gene sequence was correct.

[0062] Scale-up Culture: Inoculate the correctly sequenced bacterial culture into a 50 mL centrifuge tube containing 25 mL LB liquid medium (containing 25 μL ampicillin, 1:1000 ratio). Incubate at 37°C with shaking at 220 rpm for 12–16 hours.

[0063] Plasmid Extraction: Plasmids were extracted using the endotoxin-free plasmid extraction kit (Tiangen, catalog number DP118), and the concentration and purity (A260 / A280) were determined to ensure that the transfection requirements were met.

[0064] In vitro expression of antibodies with superior structural designs was achieved using a eukaryotic expression system: The four plasmids prepared above were transfected into HEK293F cells using the Expi293 expression system. Cell preparation: HEK293F cells were revived and cultured in suspension using Expi293™ Expression Medium. Cell density was adjusted the day before transfection to ensure that cells were in logarithmic growth phase and viability >95% on the day of transfection. Transfection complex preparation: The four plasmids were mixed according to the CrossMab design ratio (e.g., heavy chain A: light chain A: heavy chain B: light chain B = 1:1:1:1 or other optimized ratios). The plasmid DNA was mixed with the transfection reagent Lipo293F™ Plus and incubated at room temperature to form the transfection complex.

[0065] Transfection and Culture: The transfection complex was added dropwise to the cell suspension, resulting in a final transfection volume of 100 mL. Cells were placed in suspension culture at 37°C in an 8% CO2 shaker incubator. Cells were cultured for 7 days, during which time cell status was observed. The cell suspension after 7 days of culture was collected and transferred to centrifuge tubes. The cells were centrifuged at 14,000 ×g for 20 minutes, and the supernatant was carefully collected. This centrifugation step was repeated once (14,000 ×g, 20 minutes) to thoroughly remove cell debris. The collected clear supernatant was diluted 1:1 with PBS buffer (e.g., 100 mL supernatant + 100 mL PBS, final volume 200 mL) and stored in a 500 mL graduated cylinder or beaker. This step aims to adjust the pH and ionic strength of the sample to facilitate Protein G binding.

[0066] Next, the target antibody was purified using a Protein G affinity chromatography column: Protein G Agarose (FastFlow, pre-packed column) was used. The column was mounted on the purification instrument or peristaltic pump system. The tubing was rinsed: The tubing was run through with 20% ethanol to remove air bubbles, followed by column equilibration and tubing cleaning with PBS. The instrument flow rate was set, and 200 mL of the diluted sample was loaded onto the Protein G column to ensure complete antibody-to-packing ratio. After loading, the column was rinsed with PBS buffer until the baseline stabilized, removing any unbound proteins.

[0067] Elution and Neutralization: Prepare collection tubes (e.g., 1.5 mL centrifuge tubes), pre-adding 200 μL of Tris-HCl (pH 9.0) to each tube as a neutralization solution. Elute using citric acid elution buffer (pH 2.7-3.0). Collect the elution peaks in aliquots (usually 5-6 tubes). According to the UV absorption peak diagram (UV280), the target antibody is usually concentrated mainly in the second tube. Immediately shake the collection tubes to neutralize the acidic elution buffer with the pre-added Tris-HCl to prevent antibody aggregation or inactivation. Transfer the purified antibody eluent to a dialysis bag or dialysis clamp. Place in 4 L PBS buffer (or a specific formulation buffer) and dialyze with magnetic stirring at 4°C for at least 10 hours to completely remove citric acid and Tris components. Recover the dialyzed sample and filter it through a 0.22 μm filter membrane for sterilization to remove potential microorganisms and fine precipitates. After determining the protein concentration, aliquot as needed. Store long-term at -20°C. Finally, purity was analyzed by SDS-PAGE electrophoresis and SEC-HPLC.

[0068] Take the CrossMab double antibody sample, prepare the protein sample with NuPAGE reagent, denature at 70℃ for 5 minutes, and then load 8 μg of protein into each lane onto a 4–20% Tris-Glycine gel (EpiZyme); electrophoresis at 125V for about 90 minutes, and observe the bands after staining with Simply Blue safe dye: the target bands such as CD163H, CD163L, and CD16L are clearly visible in lane R (reduced form), and there are no obvious degradation product bands, which initially indicates that the sample purity is good. An Agilent 1260 system equipped with a TSKgel G3000SWxl column (7.8 mm × 300 mm, 5 μm; Tosoh) was used. The mobile phase was 0.1 M phosphate buffer (containing 0.15 M NaCl, pH 7.0) (filtered and degassed through a 0.22 μm filter). At 30 °C and a flow rate of 0.5 mL / min, 20 μL of a 1 mg / mL CrossMab antibiotic sample was injected, and the elution signals at 214 nm and 280 nm wavelengths were monitored simultaneously. Figure 4 As shown, the results indicate that only a single elution peak appeared in both the 214nm and 280nm channels. The calculated percentage of the monomer peak area to the total peak area was 97.8%, which quantitatively verifies that the purity of the CrossMab bispecific antibody sample meets the experimental requirements.

[0069] Step 5: Verify the biological characteristics of the antibody through multi-dimensional experiments. The results are as follows: Figure 5 As shown: Functional assays were performed using ELISA, flow cytometry (FCM), and confocal microscopy. Enzyme-linked immunosorbent assay (ELISA) was used to determine the binding affinity of antibodies to M1 / M2 type AMs specific markers. High-binding-affinity ELISA plates were coated with CD16A and CD163A at a concentration of 2 μg / mL and incubated overnight at 4°C. After washing with PBST and blocking with 5% BSA, serially diluted CrossMab antibiotics (with CD16 and CD163 monoclonal antibodies as controls) were added and incubated. Absorbance at 450 nm was measured using TMB colorimetry, and a dose curve was fitted (n=3). In the CD16 binding assay, the EC50 of the CrossMab antibiotics... 50 The binding affinity was 0.9 nM, with no significant difference compared to CD16 monoclonal antibody; in the CD163 binding assay, the EC50 of CrossMab bispecific antibody was [missing value]. 50The concentration was 0.8 nM, consistent with the binding activity of the CD163 monoclonal antibody, confirming that the bispecific antibody can simultaneously and efficiently bind to both CD16 and CD163 antigens. A macrophage polarization model was established and identified using flow cytometry. Mouse alveolar macrophage polarization was induced by cytokine stimulation, and CD11b was identified by flow cytometry after LPS+IFN-γ stimulation. + CD68 + CD86 + The percentage of positive cells reached 47.2±1.33%. After stimulation with IL-4 and IL-13, CD11b... + CD68 + CD206 + The percentage of positive cells was 55.4±1.76%, successfully constructing the M1 / M2 polarization model. Furthermore, CrossMab bispecific antibodies (with CD16 and CD163 monoclonal antibodies as controls) were co-incubated with polarized M1 / M2 macrophages. Flow cytometry analysis after APC-Cyanine7 conjugation with anti-human IgG Fc antibody labeling showed that the fluorescence signal distribution in the bispecific antibody group was significantly stronger than that in the monoclonal antibody group and the isotype control. Quantitative results showed that the binding rate of the bispecific antibody group to macrophages was 47.8±3.1%, significantly higher than that in the CD16 monoclonal antibody group (33.6±1.2%) and the CD163 monoclonal antibody group (26.3±0.8%), with a statistically significant difference of P<0.001, confirming that the bispecific antibody has good dual-targeting properties.

[0070] The antibody sequence includes:

[0071] Heavy chain gene sequence: signal peptide-VH-CH (Human IgG1 Mutation)

[0072] mAb-1-H1 includes:

[0073] SEQ ID NO.1:

[0074] GCGGCCGCAAACTACAAGACAGACTTGCAAAAGAAGGCATGCACAGCTCAGCACTGCTCTGTTGCCTGGTCCTCCTGACTGGGGTGAGGGCC;

[0075] SEQ ID NO.2:

[0076]

[0077] TGATTCTAGA

[0078] The amino acid sequence includes:

[0079] SEQ ID NO.3: MHSSALLCCLVLLTGVRA;

[0080] SEQ ID NO.4:

[0081] QVQLVQSGAEVKKPGESLKVSCKASGYTFTSYYMHWVRQAPGQGLEWMGIINPSGGSTSYAQKFQGRVTMTRDTSTSTVYMELSSLRSEDTAVYYCARGSAYYYDFADYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLWCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK.

[0082] Heavy chain gene sequence: signal peptide - VH - CL - Fc (Human IgG1 Mutation)

[0083] mAb - 1 - H2 includes

[0084] SEQ ID NO.5:

[0085] GCGGCCGCAAACTACAAGACAGACTTGCAAAAGAAGGCATGCACAGCTCAGCACTGCTCTGTTGCCTGGTCCTCCTGACTGGGGTGAGGGCC;

[0086] SEQ ID NO.6:

[0087]

[0088] TGATTCTAGA

[0089] The amino acid sequences include:

[0090] SEQ ID NO.7: MHSSALLCCLVLLTGVRA;

[0091] SEQ ID NO.8:

[0092] EVQLVESGGGVVQPGRSLRLSCAASGFTFSSYAMHWVRQAPGKGLEWVAVISYDGSNKYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCARENVRPYYDFWSGYYSEYYYYGMDVWGQGTTVTVSSASVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGECDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLSCAVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLVSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK.

[0093] Light chain gene sequence: Signal peptide - VL - CL (Human Lambda1)

[0094] mAb - 1 - L1 includes:

[0095] SEQ ID NO.9:

[0096] GCGGCCGCAAACTACAAGACAGACTTGCAAAAGAAGGCATGCACAGCTCAGCACTGCTCTGTTGCCTGGTCCTCCTGACTGGGGTGAGGGCC;

[0097] SEQ ID NO.10:

[0098] TCTTACGTGCTCACCCAGCCCTCCTCCGTCTCTGTTGCTCCTGGTCAGACCGCCACCATCTCCTGTGGAGGTCACAACATCGGCTCCAAGAACGTCCACTGGTACCAGCAGCGGCCCGGACAGTCTCCTGTTTTGGTCATCTACCAGGACAATAAGAGGCCCTCCGGCATCCCCGAGCGTTTTTCTGGATCTAACTCCGGTAACACCGCCACCCTCACTATCTCCGGCACCCAGGCTATGGACGAAGCCGATTACTACTGCCAGGTGTGGGACAACTACTCCGTGCTGTTCGGGGGCGGTACCAAGTTGACCGTCCTGGGTCAGCCCAAGGCCAACCCCACTGTCACTCTGTTCCCGCCCTCCTCTGAGGAGCTCCAAGCCAACAAGGCCACACTAGTGTGTCTGATCAGTGACTTCTACCCGGGAGCTGTGACAGTGGCCTGGAAGGCAGATGGCAGCCCCGTCAAGGCGGGAGTGGAGACCACCAAACCCTCCAAACAGAGCAACAACAAGTACGCGGCCAGCAGCTACCTGAGCCTGACGCCCGAGCAGTGGAAGTCCCACAGAAGCTACAGCTGCCAGGTCACGCATGAAGGGAGCACCGTGGAGAAGACAGTGGCCCCTACAGAATGTTCA;

[0099] TGATTCTAGA

[0100] The amino acid sequence includes:

[0101] SEQ ID NO.11: MHSSALLCCLVLLTGVRA;

[0102] SEQ ID NO.12:

[0103] SYVLTQPSSVSVAPGQTATISCGGHNIGSKNVHWYQQRPGQSPVLVIYQDNKRPSGIPERFSGSNSGNTATLTISGTQAMDEADYYCQVWDNYSVLFGGGTKLTVLGQPKANPTVTLFPPSSEELQANKATLVCLISDFYPGAVTVAWKADGSPVKAGVETTKPSKQSNNKYAASSYLSLTPEQWKSHRSYSCQVTHEGSTVEKTVAPTECS。

[0104] Light chain gene sequence: Signal peptide - VL - CH1 (Human IgG1)

[0105] mAb - 1 - L2 includes:

[0106] SEQ ID NO.13:

[0107] GCGGCCGCAAACTACAAGACAGACTTGCAAAAGAAGGCATGCACAGCTCAGCACTGCTCTGTTGCCTGGTCCTCCTGACTGGGGTGAGGGCC;

[0108] SEQ ID NO.14:

[0109] GATATCCAGATGACCCAGTCCCCCTCCTCCCTGTCTGCTTCTGTGGGAGACAGAGTGACCATTACCTGTAGAGCCTCCCAGTCCATTTCCTCCTACCTGAACTGGTACCAGCAGAAGCCCGGCAAGGCCCCTAAACTGCTCATCTACGCCGCCTCTAGTCTGCAGTCCGGCGTTCCTTCCCGCTTTTCCGGAAGCGGATCTGGCACCGACTTCACCCTCACCATCTCCTCCCTGCAGCCCGAAGATTTTGCTACCTACTACTGCCAGCAGTCCTACTCCACCCCCCGAGGAACATTCGGCCAGGGAACAAAAGTTGAGATCAAATCTTCCGCTAGCACCAAGGGCCCATCGGTCTTCCCCCTGGCACCCTCCTCCAAGAGCACCTCTGGGGGCACAGCGGCCCTGGGCTGCCTGGTCAAGGACTACTTCCCCGAACCGGTGACGGTGTCGTGGAACTCAGGCGCCCTGACCAGCGGCGTGCACACCTTCCCGGCTGTCCTACAGTCCTCAGGACTCTACTCCCTCAGCAGCGTGGTGACCGTGCCCTCCAGCAGCTTGGGCACCCAGACCTACATCTGCAACGTGAATCACAAGCCCAGCAACACCAAGGTGGACAAGAAAGTTGAGCCCAAATCTTGT;

[0110] TGATTCTAGA

[0111] The amino acid sequence includes:

[0112] SEQ ID NO.15: MHSSALLCCLVLLTGVRA;

[0113] SEQ ID NO.16:

[0114] DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSRFSGSGSGTDFTLTISSLQPEDFATYYCQQSYSTPRGTFGQGTKVEIKSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSC.

Claims

1. A bispecific antibody targeting heterogeneous macrophages, characterized in that: Using bevacizumab as the parental antibody backbone, the antibody arm targeting CD16 is located on the N-terminal side, and the antibody arm targeting CD163 is located on the C-terminal side. A "knob" mutation (K322A, T366W) is introduced into the CH3 domain of the heavy chain targeting CD16, while three complementary "hole" mutations (T366S, L368A, Y407V) are introduced into the CH3 domain of the heavy chain targeting CD163. Cysteine ​​residues are inserted at position S354 of the "knob chain" and position Y349 of the "hole chain". The variable regions (V regions) of CD16 and CD163 are linked in series through a (G4S)3 flexible linker and then fused with their respective constant regions (C regions).

2. The bispecific antibody targeting heterogeneous macrophages according to claim 1, characterized in that: LALA mutations (L254A, L255A) were introduced into the CH2 domain of the Fc region.

3. The method for designing a bispecific antibody targeting heterogeneous macrophages as described in claim 1 or 2, characterized in that: Screening for M1 / M2 type AMs-specific surface markers: Differential gene expression analysis was performed on single-cell sequencing datasets of AMs from public databases to screen CD16 and CD163 as the most important targets; In the bispecific antibody architecture constructed using the Kih / CrossMab-LALA technology, bevacizumab was selected as the parental antibody backbone. The antibody arms targeting CD16 were all located on the N-terminal side, while the antibody arms targeting CD163 were located on the C-terminal side. The "Knobs-into-Holes" technique was used to exchange the heavy chain CH1 domain and the light chain CL domain targeting CD163. A dual-resistance architecture is constructed using DVD-Ig technology: the variable regions (V regions) of the target CD16 and CD163 are connected in series through (G4S)3 flexible connectors, and then fused with their respective constant regions (C regions).

4. The method for designing a bispecific antibody targeting heterogeneous macrophages according to claim 3, characterized in that: The specific method of exchanging the heavy chain CH1 domain and light chain CL domain targeting CD163 involves introducing a "knob" mutation (K322A, T366W) into the heavy chain CH3 domain targeting CD16, and simultaneously introducing three complementary "hole" mutations (T366S, L368A, Y407V) into the heavy chain CH3 domain targeting CD163.

5. The method for designing a bispecific antibody targeting heterogeneous macrophages according to claim 4, characterized in that: Cysteine ​​residues are inserted at position S354 of the "knob chain" and position Y349 of the "hole chain".

6. The method for designing a bispecific antibody targeting heterogeneous macrophages according to claim 3, characterized in that: The sequence of the (G4S)3 flexible connector is: GGGGSGGGGSGGGGS.

7. The method for preparing the bispecific antibody targeting heterogeneous macrophages as described in claim 1 or 2, characterized in that, Includes the following steps: Transformation: Take 50 μL of competent E. coli, add 2 μg of target plasmid (corresponding to the four strands of vector), and incubate on ice for 30 minutes; Resuscitation: Add 500 μL of preheated LB medium to the transformed bacterial culture and incubate at 37°C in a shaker for 1 hour; Screening and sequencing: 100 μL of revived bacterial culture was spread on LB agar plates containing ampicillin and incubated upside down at 37°C overnight. The next day, single colonies were picked and inoculated into 1.5 mL of LB liquid medium containing antibiotics. The culture was shaken at 37°C, and plasmids were extracted and sent for sequencing. The sequences were compared to confirm that the gene sequences were correct. Expanded culture: Inoculate the correctly sequenced bacterial culture into 25 mL LB liquid medium and incubate at 37°C in a shaker for 12-16 hours; Plasmid extraction: Plasmids were extracted using an endotoxin-free plasmid extraction kit; Using the Expi293 expression system, the four plasmids prepared above were transfected into HEK293F cells for expression. The transfection complex was added dropwise to the cell suspension, and the cells were placed in a 37°C, 8% CO2 shaker incubator for suspension culture.

8. The method for preparing the bispecific antibody targeting heterogeneous macrophages according to claim 7, characterized in that: The target plasmid includes at least one set of nucleotide sequence fragments: SEQ ID NO.1 and SEQ ID NO.2, SEQ ID NO.5 and SEQ ID NO.6, SEQ ID NO.9 and SEQ ID NO.10, SEQ ID NO.13 and SEQ ID NO.14.

Images