Mini-protein antagonists targeting il-4rα, pharmaceutical compositions thereof, and uses thereof

Abstract

The application belongs to the field of protein drugs, and particularly relates to a mini-protein antagonist targeting IL-4R alpha, a pharmaceutical composition thereof and application thereof. The mini-protein of the application comprises an amino acid sequence selected from SEQ ID No. 21-40, 57-58. The high-affinity IL-4R alpha mini-protein antagonist of the application exhibits good specificity and thermal stability, and still maintains its biological activity unchanged after heating at 95 DEG C. In-vitro human IL-4 / IL-13 signal blocking experiments prove that the mini-protein of the application exhibits a blocking effect comparable to dupilumab on the signals of the two cytokines, human IL-4 / IL-13.

Description

Technical Field

[0001] This application relates to the field of protein drugs, specifically to a mini protein antagonist targeting IL-4Rα, pharmaceutical compositions thereof, and their applications. Background Technology

[0002] IL-4Rα is a key regulatory receptor for type 2 inflammation. It transmits signals from both IL-4 and IL-13 cytokines by binding to IL-13Rα1 or γc chains, inhibiting the activation of the downstream JAK1-STAT6 signaling pathway. IL-4Rα is currently the most successful gold standard target in the treatment of allergic diseases. Its targeted drug, dupilumab, through a dual IL-4 / IL-13 blocking mechanism, has ushered in a new era of precision treatment for type 2 inflammation, marking a milestone in clinical significance and consistently ranking first in global sales of drugs for allergic and autoimmune diseases. As of 2025, dupilumab has been approved for marketing in more than 60 countries worldwide (including China), with annual sales exceeding US$18 billion. Approved indications include eight diseases: atopic dermatitis, moderate to severe asthma, chronic sinusitis with nasal polyps, eosinophilic esophagitis, nodular prurigo, chronic obstructive pulmonary disease, chronic spontaneous urticaria, and bullous pemphigoid, and the scope of indications continues to expand. In addition to Dupilumab, the IL-4Rα-targeting antibody drug Stapokibatra has also been approved for marketing in China. Furthermore, more than 20 other IL-4Rα-targeting antibody drugs are in clinical or preclinical research stages, fully demonstrating the effectiveness of IL-4Rα-targeting antibody therapy. The central role of 4Rα in the treatment of allergic diseases.

[0003] Although IL-4Rα antagonist antibodies are highly effective, their long-term accessibility to patients with allergies is limited by the need for long-term injections for chronic diseases, the approximately bi-weekly injection cycle of traditional antibody drugs, and their high cost and production expenses. Compared to traditional antibodies, de novo-designed mini-binding proteins offer significant advantages, including easy expression via E. coli, high expression levels, strong thermostability, high solubility, and strong tissue penetration. Furthermore, mini-binding proteins possess more flexible engineering properties, allowing for easier fusion and property modification. De novo design of mini-proteins has resulted in binding proteins targeting immune components such as TNFR, IL1R, IL-2Rβ, IL6R, IL17R, C9, EGFR, and TGFβRII, but no reports have been made targeting IL-4Rα, a key therapeutic target for allergic diseases. Antibody drugs, such as Dupilumab, can competitively bind to IL-4Rα with extremely high affinity, achieving excellent blocking effects at low antibody concentrations, while also possessing monomeric properties and excellent solubility. Whether de novo designed mini-binding proteins can achieve effects comparable to monoclonal antibody drugs while maintaining their advantages is a key issue in evaluating their drugability. Summary of the Invention

[0004] The technical problem to be solved by this application is to provide a mini protein that can competitively bind to IL-4Rα with human IL-4 and IL-13 with extremely high affinity, which can provide effects comparable to monoclonal antibody drugs.

[0005] On the one hand, the present invention provides a mini protein comprising an amino acid sequence selected from SEQ ID No. 21-40, 57-58.

[0006] Preferably, the mini protein comprises the amino acid sequence of SEQ ID No. 28.

[0007] Preferably, the amino acid sequence of the mini protein is selected from SEQ ID No. 1-20, 50 and 54.

[0008] More preferably, the amino acid sequence of the mini protein is SEQ ID No. 8.

[0009] On the other hand, the present invention provides a fusion protein obtained by fusing the above-mentioned mini protein with the HSA domain ABD035 of serum albumin.

[0010] In particular, fusion with the aforementioned ABD035 protein can prolong the in vivo half-life of the mini protein.

[0011] More preferably, the amino acid sequence of the fusion protein is selected from SEQ ID No. 55 and SEQ ID No. 56.

[0012] In another aspect, the present invention provides a pharmaceutical composition comprising the aforementioned mini protein or the aforementioned fusion protein, and pharmaceutically acceptable excipients.

[0013] In another aspect, the present invention provides the use of the above-mentioned mini protein, the above-mentioned fusion protein, or the above-mentioned pharmaceutical composition in the preparation of IL-4Rα-targeting antagonists.

[0014] In another aspect, the present invention provides the use of the above-mentioned mini protein, the above-mentioned fusion protein, or the above-mentioned pharmaceutical composition in the preparation of a medicament for treating allergic diseases.

[0015] Preferably, the allergic disease is selected from atopic dermatitis, asthma, chronic sinusitis with nasal polyps, allergic rhinitis, hay fever, and eosinophilic esophagitis.

[0016] Beneficial effects This application utilizes a deep learning-based de novo design approach to design and screen high-affinity IL-4Rα mini-protein antagonists through large-scale cluster computing. Furthermore, the binding affinity was optimized from 22.1 nM to 569 pM (Binder-82-O32) via partial diffusion, resulting in a high-affinity human IL-4Rα (hIL-4Rα) mini-protein binding molecule.

[0017] The IL-4Rα mini protein antagonist designed in this application exhibits good specificity and thermal stability, and retains its biological activity even after heating at 95°C.

[0018] In vitro human IL-4 / IL-13 (hIL-4 / hIL-13) signal blocking experiments confirmed that these novel mini-protein antagonists of this application exhibit a good blocking effect on the signaling of both hIL-4 and hIL-13 cytokines.

[0019] Furthermore, this application binds the mini protein antagonist to the HSA-binding domain ABD035 of serum albumin, successfully extending the half-life from 2.706 hours to 60.6 hours, thus solving the problem of excessively rapid metabolism of mini proteins. Attached Figure Description

[0020] Figure 1This diagram illustrates the design, synthesis, and screening of the IL-4Rα mini-protein antagonists in this application, as well as their application in extending the half-life in vivo. The diagram schematically shows the research process of this application: firstly, a series of IL-4Rα mini-protein antagonists were obtained through initial screening; then, the affinity of the screened mini-protein antagonists for IL-4Rα was enhanced using partial diffusion technology; next, the crystal structure and specificity of the mini-protein antagonists were analyzed and compared with monoclonal antibodies; finally, the half-life was extended by fusing the HSA-binding domain ABD035.

[0021] Figure 2 Computational design and screening of high-affinity IL-4Rα mini-protein antagonists are shown. (A) Mechanism of human IL-4Rα (hIL-4Rα) mini-protein antagonists blocking the human IL-4 / IL-13 (hIL-4 / hIL-13) signaling pathway. (B) Binding interface of hIL-4 (red) and hIL-4Rα (dark blue). A magnified view highlights hotspot residues (orange) used for conjugate design. (C) Schematic diagram of the mini-protein antagonists obtained from preliminary BLI screening. Overlay of the parental conjugate Binder-82 (light blue) with partially diffusion-optimized Binder-82-O32 (red); magnified view shows the optimized interface adjustments. (D) Preliminary BLI screening results of the interaction between candidate mini-protein inhibitors and hIL-4Rα; six candidates with superior performance are shown in yellow. (E) Schematic diagram of the signal reporter mechanism in the hIL-4 / hIL-13 Reporter 293 cell line. (F) Evaluation of the blocking effect of six mini-protein antagonists on IL-4-induced STAT6 signaling pathway activation. (G) Evaluation of hIL-4-induced STAT6 pathway inhibition by 150 partially diffused optimized mini-protein antagonists (500 nM) using the hIL-4 / hIL-13 Reporter 293 cell line. Values ​​with normalized luciferase activity <0.1 are highlighted in yellow. (H) Further evaluation of the blocking effect of optimized candidates on hIL-4-induced STAT6 signaling at a concentration of 100 nM of mini-protein inhibitor. (I) Final comparison of six candidates at a concentration of 20 nM of mini-protein inhibitor to evaluate their blocking effect on hIL-4-induced STAT6 signaling. (J) Affinity between mini-proteins Binder-82, Binder-82-O32 and hIL-4Rα was determined using BLI. Data in Figures F, H, and I are presented as mean ± standard deviation, n = 3 biologically independent samples. ns indicates no statistical significance, P<0.05, P<0.01, P<0.001, ****P<0.0001.

[0022] Figure 3The binding specificity of Binder-82-O32 is demonstrated and compared with Dupilumab. (A) The superimposed structure of the Binder-82-O32-M2 crystal structure (orange) and the predicted Binder-82-O32 (red) / hIL-4Rα (dark blue) complex model. A magnified view shows three key interacting residues at the Binder-82-O32-hIL-4Rα binding interface. (B) The blocking effect of the crystalline construct Binder-82-O32-M2 on the hIL-4-induced STAT6 signaling pathway. (C) FACS data and mean fluorescence intensity (MFI) plots showing the binding ability of mini-protein antagonists with arginine mutations at different key interaction sites to hIL-4Rα-positive cells. (D) The blocking effect of Binder-82-O32 variants with arginine mutations at different key interaction sites on the hIL-4 or hIL-13-induced STAT6 signaling pathway. (E, F) Dose-dependent inhibition of the hIL-4 / hIL-13-induced STAT6 signaling pathway by mini protein antagonists and Dupilumab. IC50 in Figures E and F. 50 The values ​​were calculated based on the corresponding fitted curves. (G) Circular dichroism chromatograms of Binder-82-O32 under different conditions (gray, 25°C; red, 95°C; orange dashed line, 95°C heat-treated and then cooled to 25°C). (H) Comparison of the bioactivity of Binder-82-O32 before and after heat treatment with Dupilumab as a reference standard. (I) Comparison of the blocking activities of Binder-82-O32 and ABD035-Binder-82-O32 at gradient dilution concentrations. (J) Schematic diagram of the structure of ABD035-Binder-82-O32. (K) Changes in in vivo drug concentrations of Cy5-labeled Binder-82-O32 and ABD035-Binder-82-O32 in 6 Wistar rats (n=3). Data in Figures B, C, D, E, F, H, I, and K are expressed as mean ± standard deviation, n=3 biologically independent samples. ns indicates no statistical significance, P<0.05, P<0.01, P<0.001, ****P<0.0001.

[0023] Figure 4The calculated scores for the design of mini-proteins binding to hIL-4Rα are shown. AlphaFold2 predicted scores for the first (A), second (B), and third (C) rounds of screening. Scores meeting the threshold criteria of pae_interaction < 10 and plddt_binder > 90 are highlighted in red. (D) AlphaFold3 predicted scores for mini-protein antagonists with pae_interaction < 6 in the three rounds of screening. (E) AlphaFold3 predicted scores for mini-protein antagonists derived from Binder-36 and Binder-82 with pae_interaction < 5.4 after partial diffusion optimization. Short binders are represented as short mini-proteins with a length of 60–115 residues, and long binders are represented as long mini-protein backbones with a length of 115–140 residues.

[0024] Figure 5 Preliminary BLI screening results for the interaction between candidate mini-protein antagonists and hIL-4Rα are shown. BLI data on the binding of mini-protein antagonists to hIL-4Rα in the initial screening are presented.

[0025] Figure 6 The binding kinetics between Dupilumab and hIL-4Rα are shown.

[0026] Figure 7 The design and screening data for mini-protein antagonists targeting mIL-4Rα are shown. (A) BLI data show that four mini-proteins that bind to hIL-4Rα do not interact with mIL-4Rα. (B) AlphaFold2 prediction scores for mini-protein antagonists targeting mIL-4Rα. Scores meeting the threshold criteria of pae_interaction < 10 and plddt_binder > 90 are highlighted in red. (C) Statistical graph of response values ​​in Figure D. (D) BLI screening results of the interaction between 60 candidate mini-protein antagonists and mIL-4Rα. (E) Structural schematic diagrams of Binder-17, Binder-23, and Binder-25 in Figure (D) selected by BLI.

[0027] Figure 8The image shows size exclusion chromatography (SEC) analysis of the complexes of Binder-82-O32 and its alanine mutant Binder-82-O32-M2 with hIL-4Rα. (A) SEC chromatograms show that both Binder-82-O32 and its alanine mutant Binder-82-O32-M2 can form complexes with hIL-4Rα. (B) Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of the fractions collected from each elution peak in Figure A. (C) Size exclusion chromatography-multi-angle light scattering characterization of Binder-82-O32.

[0028] Figure 9 SDS-PAGE analysis of hIL-4Rα mini-protein antagonists is shown. (A) SDS-PAGE analysis of mini-protein antagonists (5 μg each) after Ni-NTA affinity purification. (B) SDS-PAGE analysis of Binder-82-O32-43C-Cy5 and ABD035-Binder-82-O32-Cy5 under oxidizing (copper sulfate) and non-oxidizing conditions. Figure A gels were stained with Coomassie Brilliant Blue; Figure B gels were unstained. Detailed Implementation

[0029] The technical solutions of this application are described in detail below through specific embodiments to enable those skilled in the art to better understand this application; however, the provision of these embodiments is not intended to limit the scope of this application.

[0030] the term: In this application, the terms "mini protein" and "mini protein antagonist" are used interchangeably, both referring to protein molecules that target IL-4Rα and antagonize its interaction with IL-4 / IL-13.

[0031] In the following text, the term "pae_interaction" used in mini-binding protein screening refers to the Predicted Aligned Error for the interaction interface, which is a metric specifically used to assess the reliability of inter-chain interaction interfaces.

[0032] The research process of this application will be described in detail below, such as Figure 1As shown, a series of IL-4Rα mini protein antagonists were first obtained through primary screening. Then, the affinity of the primary screened mini protein antagonists for IL-4Rα was enhanced by partial diffusion technology. Next, the crystal structure and specificity of the mini protein antagonists were analyzed and compared with monoclonal antibodies. Finally, the half-life was extended by fusing the HSA binding domain ABD035.

[0033] Example 1: Generating a mini protein antagonist targeting IL-4Rα using computational protein design via RFdiffusion. Blocking IL-4Rα signaling requires inhibiting its interaction with IL-4 / IL-13. Figure 2 A).

[0034] This application focuses on designing a cytokine-binding interface for mini-protein inhibitors to bind to IL-4Rα. Figure 2 B). Figure 4 The computational score for the design of the mini-protein binding to hIL-4Rα is shown. In the first round of computational design for hIL-4Rα, four hydrophobic residues, L64, F66, V94, and Y152, were selected as hotspot residues, and 16,000 protein backbones (50-115 residues) were generated using RFdiffusion. Figure 2 B). After ProteinMPNN sequence design and AlphaFold2 complex prediction, screening was performed using pae_interaction<10 and plddt_binder>90 as conditions, with a success rate of 1.85%. Figure 4 A). Analysis showed that the successfully binding protein had a broader interaction interface than cytokines, exceeding the pre-defined hotspot range (A). Figure 2 B and Figure 2 C). It is worth noting that the double helix structure accounts for a high proportion of the successfully designed mini proteins, while this type of backbone has performed poorly in previous studies, suggesting that the initial backbone length may be insufficient.

[0035] Accordingly, hotspot residues Y179 and L180 were added in the second round of design. Figure 2 (B) This generated 16,000 short mini-protein backbones with lengths of 60-115 residues and long mini-protein backbones with lengths of 115-140 residues. The screening success rates increased to 2.44% and 5.13%, respectively. Figure 4 B and 4C). Mini protein antagonists satisfying pae_interaction < 6 were further evaluated using the AlphaFold3 score. Results showed that mini protein antagonists with longer backbones exhibited better success rates and ipTM values ​​(B and 4C). Figure 4D). Based on the combined results of the two rounds of calculations, a total of 100 mini protein antagonists with high ipTM scores were selected for further experimental screening.

[0036] For mIL-4Rα, this application uses the same calculation process for screening. Figure 7 This study presents the design and screening data for mini-protein antagonists targeting mIL-4Rα.

[0037] The above results show that the design success rate of mini protein antagonists targeting mIL-4Rα was 1.31%. Figure 7 (B) Significant differences exist between human and murine receptors: successfully binding proteins of mIL-4Rα are mainly concentrated at the cytokine binding interface and have relatively short backbones, suggesting that cross-species activity may be limited. Based on this, the top 60 candidate molecules were selected for further experimental validation.

[0038] Example 2: Initial screening of mini protein antagonists targeting IL-4Rα The coding sequence of the mini protein antagonist was optimized by the expression codon of E. coli, synthesized by Universe Gene Technology or GENCEFE Biotech, and constructed between the NcoI and XhoI restriction sites of the pET-28a vector. The plasmid was then transformed into E. coli BL21(DE3) for expression. The specific expression procedure is as follows: Transformed single colonies were inoculated into 100 mL LB medium containing 50 μg / mL kanamycin and cultured at 37°C with shaking at 220 rpm. When the OD600 reached 0.6–0.8, 0.3 mM IPTG was added, and protein expression was induced at 24°C for 12–16 hours. After expression, the cells were collected by centrifugation at 10,000 × g, resuspended in PBS, and then sonicated to lyse. The lysate was centrifuged at 12,000 × g for 20 minutes to remove insoluble impurities. The supernatant was loaded onto a pre-equilibrated Ni-NTA affinity chromatography column for purification. Non-target proteins were removed by washing with PBS solutions containing 10, 20, and 30 mM imidazole sequentially. Finally, the target protein was eluted with PBS containing 300 mM imidazole (pH 7.4). Except for one candidate molecule that failed to be expressed normally during the initial screening of hIL-4Rα, all other candidate proteins were successfully expressed.

[0039] After expression and His-tag purification in *E. coli*, biological layer interference (BLI) was used for screening, and six mini proteins that bind to hIL-4Rα and three that bind to mIL-4Rα were identified as having high binding affinity. Figure 2 D and Figure 7 C), Figure 7 E shows a schematic diagram of the structures of the three mini proteins, Binder-17, Binder-23, and Binder-25.

[0040] The amino acid sequences of the six hIL-4Rα-binding mini-protein antagonists are SEQ ID No. 41, SEQ ID No. 42, SEQ ID No. 43, SEQ ID No. 44, SEQ ID No. 45, and SEQ ID No. 46, respectively; the amino acid sequences of the three mIL-4Rα-binding mini-protein antagonists are SEQ ID No. 47, SEQ ID No. 48, and SEQ ID No. 49, respectively, as shown in Table 1 below.

[0041] Table 1: Amino acid sequences of IL-4Rα mini-binding protein antagonists and their corresponding BLI response values

[0042] In the above sequences, the starting amino acid M is the first part, and the ending part (ggsHHHHHH) indicates linker+6*His. The first 6 in the table are mini proteins that can bind to hIL-4Rα with high affinity, and the last 3 are mini proteins that can bind to mIL-4Rα with high affinity.

[0043] The above BLI experiments were performed using the OCTET RED96E system (ForteBio), and the data were processed using the accompanying software (Fortebio Data Analysis 12.0.1.2). All experiments were performed according to the standard procedure including baseline equilibration, hIL-4Rα immobilization, second baseline equilibration, binding, and dissociation, with the baseline step set to 60 seconds. The Fc-tagged hIL-4Rα protein (Novoprotein, CS38; ACRO Biosystems, ILR-M52H1) was diluted to a final concentration of 5 μg / mL as the immobilized protein and immobilized on a Protein A sensor (ForteBio) for 600 seconds; the binding time was set to 330 seconds, and the dissociation time was set to 300 seconds. For initial screening, the mini protein antagonist was diluted to 1 μM (… Figure 5 and Figure 6 In gradient binding experiments, the fractions are diluted to 500-1.95 nM. Figure 2 J, Figure 5 The buffer used was PBST (PBS containing 0.05% Tween-20, pH 7.4). The flow rate was 600 rpm for fixed steps and 1000 rpm for other steps. All experiments were performed at 25°C. Figure 2 J and Figure 6 The data in the model are fitted globally using a 1:1 combination. Figure 5 and Figure 7 A, Figure 7The data in D were fitted using a 1:1 model with local fitting. The fitting parameters for Binder-82 and Dupilumab were Rmax 0.3766 and R... 2 0.9964, X 2 0.2161, K d 22.18 nM and Rmax 0.4824, R 2 0.9967, X 2 10.4832, K d The concentration was less than 100 pM, and a control well was included in each independent experiment to subtract background. The Dupilumab used in this study was purchased from Sanofi.

[0044] Given that hIL-4Rα has greater potential for drug development, this application focuses on the above-mentioned six candidate molecules targeting hIL-4Rα. Figure 2 C), and evaluated its blocking effect on hIL-4 / hIL-13 signaling in the luciferase reporter gene 293 cell line stably expressing hIL-4Rα / hIL-13Rα1. The blocking experiment was as follows: hIL-4 / hIL-13 Reporter 293 cell line (Genomeditech, GM-C01511) was cultured in DMEM (Gibco) medium containing 10% FBS (Gibco), 1% penicillin-streptomycin (10,000 U / mL, Gibco), blastomycin (4 μg / mL, MCE, HY-K1054), and puromycin (0.75 μg / mL, Sangon Biotech, E607054-0001) in an incubator at 37°C, 5% CO2, and approximately 80% relative humidity; all cell lines were used within 2 months after thawing and mycoplasma contamination was tested regularly. In functional experiments, cells were cultured at a density of 1.5 × 10⁻⁶. 4 Seeds were planted at the specified density in 96-well plates 20–24 h before stimulation and incubated overnight at 37°C with 5% CO2. After removing the supernatant, serially diluted mini-protein antagonists, Dupilumab, recombinant hIL-4 (100 pM, Novoprotein, CX03), or hIL-13 (1 nM, Novoprotein, CC89) were prepared in DMEM containing 1% FBS (Gibco) and added to the corresponding wells. The experimental groups were as follows: experimental wells (mini-protein antagonists + cytokines), Dupilumab control wells (serially diluted Dupilumab + cytokines), positive control wells (cytokines only), and blank control wells (DMEM containing only 1% FBS). After incubation at 37°C for 7 hours, 100 μL of Bio-Lumi was added to each well. TMII. Firefly luciferase reporter gene assay kit (Beyotime, RG043M). After equilibration at room temperature for 5 minutes, the sample was transferred to a white opaque assay plate, and chemiluminescence was measured using a Tecan Spark microplate reader. Candidate proteins identified through BLI screening were first tested for their inhibitory activity against IL-4 signaling at concentrations of 500 and 100 nM. Results showed that Binder-82 and Binder-36 achieved >85% pathway blockade, but their inhibitory efficacy was still lower than that of Dupilumab (…). Figure 2 E and Figure 2 F). Based on this, the partial diffusion strategy is further employed in the following embodiments to optimize the affinity of Binder-82 and Binder-36.

[0045] Example 3: Enhancing the affinity of mini protein antagonists for IL-4Rα through partial diffusion In the partial diffusion optimization stage, Binder-82 and Binder-36 were used as starting templates. The process was executed under the condition that `diffuser.partial_T` = 6, 8, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 23, 25, generating 1,250 skeletons for each T value. Subsequent processing was performed using ProteinMPNN sequence design and AlphaFold2 structure prediction. Models were screened using `pae_interaction < 5.4` and `plddt_binder > 92` as thresholds, and AlphaFold3 was further used for structure scoring. Based on ipTM ranking, 81 and 69 candidate molecules were selected for Binder-82 and Binder-36, respectively. Figure 4 E) Synthesis was performed. After protein expression and purification, candidate molecules were directly screened for function using the hIL-4 / hIL-13 Reporter 293 cell line: First, a first round of screening was conducted at a concentration of 500 nM for the mini protein antagonist, yielding 20 mini protein antagonists capable of achieving complete pathway blockade (amino acid sequences are shown in Table 2 below). Figure 2 G); After reducing the concentration of mini protein antagonists to 100 nM, the six mini protein antagonists, Binder-82-O20, O32, O34, O60, O73, and O79, could still completely block cell signal transduction. Figure 2 H); further reducing the concentration of the mini protein antagonist to 20 nM, it was found that only Binder-82-O32 could still completely block the signaling pathway (H); Figure 2 I). Gradient BLI experiments showed that the affinity of Binder-82-O32 for IL-4Rα increased from 22.1 nM to 569 pM ( Figure 2The J) indicates that partial diffusion significantly enhances binding affinity. The BLI fitting parameters for Binder-82-O32 are Rmax 0.6492, R2 0.9957, X2 0.2317, and Kd 0.569 nM.

[0046] Table 2: Amino acid sequences and inhibitory activities (%) of the partially diffusion-optimized Binder-36 / 82 mini-protein antagonists

[0047] In the sequences above, the starting amino acid M is at the beginning, and the ending part (ggsHHHHHH) represents linker+6*His. The sequences after removing the N-terminal MG from each of the above sequences are SEQ ID No. 21-40.

[0048] The relative luciferase activities in the table are derived from data analysis using the formula Relative luciferase activity (%) = (Relative luciferase activity of experimental group - Relative luciferase activity of blank group) / (Relative luciferase activity of positive control group - Relative luciferase activity of blank group) after IL-4 signal blocking experiments with hIL-4 / hIL-13 Reporter 293 cells.

[0049] Example 4: Crystallization and structural analysis of IL-4Rα mini protein antagonists To assess the design accuracy, we first attempted co-crystallization of hIL-4Rα and Binder-82-O32. The extracellular domain (M26-E227) of hIL-4Rα, after codon optimization, was fused with a mouse Igκ secretion signal peptide and a (His)6 tag, inserted between the XbaI and BamHI sites of the pcDNA3.4 vector. The plasmid was sterilely filtered and transiently transfected using a mammalian expression system. After 5 days of expression, the supernatant was collected, and the (His)6-tagged hIL-4Rα was purified using Ni Smart beads (Smart-Lifesciences) according to the manufacturer's protocol. Subsequently, the purified hIL-4Rα was mixed with an excess of Binder-82-O32, incubated at room temperature for 30 minutes, and then purified by gel chromatography (SEC) using a Superdex 75 Increase 10 / 300 GL column (GE Healthcare) on an FPLC system (Union-biotech, UEV25D). IL-4Rα formed a tight complex with Binder-82-O32 ( Figure 8However, no crystals with good diffraction were obtained. Therefore, the focus was on crystal screening of the Binder-82-O32 single protein: initial screening was performed using kits (Crystal Screen, PEGRx, and Wizard Classic 1-4) purchased from Hampton Research and Rigaku, employing a two-site mosquito LCP (SPT Labtech) in a 96-well crystallization plate using a sitting drop method for initial screening and optimization. Given the strong hydrophilicity of Binder-82-O32, several surface hydrophilic amino acids outside the IL-4Rα binding interface were mutated to alanine to promote crystal growth (PMID: 40877575, 41813685), constructing mutants containing 3–4 surface alanine mutations, with the (His)6 tag directly fused to the designed sequence; the mini-protein used for crystal screening was further purified by gel chromatography and concentrated by ultrafiltration to approximately 75 mg / mL. A crystal of Binder-82-O32-M2 (amino acid sequence SEQ ID No. 50, with the N-terminal MG removed and the C-terminal His tag removed, sequence SEQ ID No. 57) was successfully obtained. The sequence is shown in Table 3. The crystallization conditions were 6% v / v isopropanol, 0.1 M sodium acetate trihydrate (pH 4.5), and 26% v / v polyethylene glycol monomethyl ether 550. Diffraction and data collection were performed using an indoor X-ray diffraction system (Rigaku) ​​equipped with Cu Kα radiation (wavelength 1.54 Å). The resolved crystal structure resolution was 3.5 Å. The structure was resolved using molecular substitution, with the design model as the search template. Molecular substitution and structural refinement were performed using the PHENIX software package and COOT. Structural visualization was performed using ChimeraX and PyMOL. Data collection and refinement statistics are detailed in Table 4 below. Structural comparison showed a root mean square deviation of 1.048 Å between the design model and the crystal structure, indicating high design accuracy. Figure 3 A). Functional experiments at 100 nM and 20 nM concentrations (IL-4 only) showed that the crystallographic construct Binder-82-O32-M2 had similar inhibitory activity to Binder-82-O32. Figure 3 B).

[0050] To verify the specificity of the binding interface, three interface residues A12, E100, and L111 in Binder-82-O32 were mutated to arginine (the amino acid sequences are shown in Table 3 below, namely SEQ ID No. 51, SEQ ID No. 52, and SEQ ID No. 53, respectively). Figure 3 A); Flow cytometry results of cell surface binding (4 μM) showed that all three mutations weakened binding to IL-4Rα-positive 293 cells ( Figure 3C); Cell signal blocking assays (4 and 0.8 nM, IL-4 and IL-13) also showed that all three mutations significantly reduced the blocking ability of Binder-82-O32 (C); Figure 3 D).

[0051] Table 3: Alanine mutant sequences for crystal screening and arginine mutant sequences for verifying binding site specificity in Binder-82.

[0052] The specific procedure for the cell surface binding assay is as follows: hIL-4 / hIL-13 Reporter 293 cells (1 × 10⁻⁶) 5 Cells were seeded in 24-well plates (number per well) and cultured overnight. Cells were then incubated with the mini-protein binding protein at 4°C for 30 minutes; an equal volume of PBS was added to the control group. After incubation, cells were washed three times with PBS to remove unbound protein. Cells were then stained with rabbit anti-6×His monoclonal antibody (Proteintech, CL647-66005) at 4°C for 30 minutes. After washing three more times, cells were resuspended in PBS and analyzed using a BD FACSAria III flow cytometer (BD Biosciences). Data were analyzed using FlowJo software (version X10.0.7r2), and binding strength was quantified as mean fluorescence intensity (MFI).

[0053] Table 4. Statistical Table of Crystal Structure Data Collection and Refinement

[0054] Example 5: Activity characterization and thermal stability evaluation of IL-4Rα mini protein antagonists Based on the above experiments, a gradient concentration signaling pathway blocking experiment was further conducted to determine and compare the half-maximal inhibitory concentrations (IC50) of Binder-82-O32 and Dupilumab on the hIL-4Rα-mediated hIL4 / hIL13 signaling pathway. 50 The inhibitory activities of Binder-82-O32 and Dupilumab were directly compared. The results showed that Binder-82-O32 and Dupilumab blocked the hIL-4 activation signal at IC50. 50 The values ​​were 0.3145 nM and 0.1175 nM, respectively, for IC blocking the hIL-13 activation signaling pathway. 50 The values ​​were 4.463 pM and 1.034 pM, respectively, indicating that the monovalent hIL-4Rα mini protein antagonist can achieve an effect comparable to that of the bivalent dupilumab. Figure 3 E and Figure 3 F).

[0055] Further evaluation of the thermal stability of Binder-82-O32: Circular dichroism spectroscopy showed that the secondary structure of Binder-82-O32 was completely recovered after heating at 95°C for 5 minutes and cooling to room temperature. Figure 3 G). The inhibitory effects of Binder-82-O32 on hIL-4 and hIL-13 signaling were tested at a concentration of 20 nM. The results showed that Binder-82-O32 completely retained its inhibitory activity, while Dupilumab completely lost its signaling pathway inhibitory activity. Figure 3 This indicates that Binder-82-O32 has a significant advantage over Dupilumab in terms of bioactive thermal stability.

[0056] In addition, Binder-82-O32 showed a high expression level ( Figure 9 The concentration reached approximately 60.5 mg / L without optimization. All gradient experiments used serial dilution methods to prepare the gradient dilution buffers. Blank wells were not treated with IL-4 or IL-13. The inhibition rate was calculated using the following formula after standardization of luciferase activity: Standardized luciferase activity = (sample luminescence value – blank luminescence value) / (positive luminescence value – blank luminescence value), luciferase inhibition rate = [1 – (sample luminescence value – blank luminescence value) / (positive luminescence value – blank luminescence value)] × 100.

[0057] Example 6: In vivo half-life extension strategy of hIL-4Rα mini protein antagonist Laboratory animals were purchased from Jinan Pengyue Laboratory Animal Breeding Co., Ltd. All animal experimental procedures strictly adhered to the protocols established by the Shandong Provincial Laboratory Monitoring Committee of China and obtained ethical approval from the Laboratory Animal Ethics Committee of Shandong Second Medical University (No.: NO.2025SDL799). Wistar rats were housed under specific pathogen-free conditions in a 12-hour light / dark cycle environment and fed standard feed; the temperature in the housing was maintained at 20–24℃, and the relative humidity was maintained at 50%–60%.

[0058] The mini protein 82-O32-43C containing a cysteine ​​mutation (SEQ ID No:54, the above sequence with the N-terminal MG removed and the C-terminal ggsHHHHHH sequence is SEQ ID No.58) was reduced by TCEP (Beyotime, ST046) and then fluorescently labeled with sulfo-Cy5 maleimide (DuoFluor, 2242791-82-6). The purified cysteine-mutated mini-proteins Binder-82-O32-43C and the fusion protein of Binder-82-O32-43C and ABD035 (ABD035-82-O32-43C SEQ ID No:55, amino acid sequence shown in Table 5 below) were mixed with TCEP at a molar ratio of 10:1 and reacted overnight at 4°C. After removing TCEP by SEC, the mini-proteins were collected, excess Cy5 dye was added, and the mixture was incubated at room temperature by rotation for 2 hours. After labeling, unbound Cy5 dye was removed by SEC purification using an FPLC system (Union-biotech, UEV25D) through a Superdex 75 Increase 10 / 300 GL column (GE Healthcare). To detect the degree of Cy5 labeling, CuSO4 with a final concentration of 2 mM was added to the sample (the control group was treated with an equal amount of PBS), and the sample was oxidized at room temperature for 30 minutes. Subsequently, 10 μg of protein sample was mixed with 5× loading buffer, heated at 95°C for 5 minutes, and then analyzed by SDS-PAGE.

[0059] Table 5: Mini-protein sequences mentioned in the in vivo half-life extension strategy of hIL-4Rα mini-protein antagonists

[0060] Note: ABD035-82-O32 in the table is a fusion protein formed by mini protein 82-O32 and ABD035 (see structural diagram). Figure 3 J) does not contain cysteine ​​mutations, while 82-O32-43C and ABD035-82-O32-43C are obtained by single-point cysteine ​​mutations at 82-O32 or ABD035-82-O32.

[0061] Blocking experiments were conducted on Binder-82-O32 and ABD035-Binder-82-O32 at gradient dilution concentrations. The results are shown below. Figure 3 I.

[0062] Subsequently, male SPF-grade Wistar rats (6–8 weeks old, n=3 per group) were subcutaneously injected with a single dose of Cy5-labeled mini-protein at a dose of 3 mg / kg. Blood samples were continuously collected via the orbital venous plexus or tail vein at 0, 0.5, 1, 1.5, 2, 2.5, 3, 4, 6, 8, 10, 12, 13, 14, 24, 32, 48, 56, 72, 80, 96, 104, 120, 128, and 192 hours after administration. Serum was collected after centrifugation, and the fluorescence signal intensity at an excitation wavelength of 630 nM and an emission wavelength of 680 nM was detected using the fluorescence module of a Tecan Spark microplate reader to construct plasma concentration-time curves. Figure 3 (K) Pharmacokinetic parameters were calculated to characterize the in vivo pharmacokinetic properties of the compound, and the half-life was calculated using nonlinear single-compartment decay fitting. Results showed that Binder-82-O32 binding to the HSA domain ABD035 of serum albumin successfully extended the half-life from 2.706 hours to 60.6 hours, providing a method to address the problem of excessively rapid metabolism of mini proteins. The normalized fluorescence intensity was calculated using the formula F. λ / F max , of which F λ F represents the fluorescence intensity at wavelength λ. max The maximum fluorescence intensity is at the emission peak.

[0063] In summary, this application successfully developed a high-performance mini protein antagonist targeting IL-4Rα through large-scale cluster computing and de novo design based on deep learning. Although Dupilumab exhibits stronger binding affinity (K... d <100pM) Figure 6 However, the broader blocking interface allows mini protein antagonists to achieve signaling pathway inhibition levels comparable to Dupilumab.

Claims

1. A mini protein, characterized in that, It contains amino acid sequences selected from SEQ ID No. 21-40, 57 and 58.

2. The mini protein according to claim 1, characterized in that, The mini protein contains the amino acid sequence of SEQ ID No.

28.

3. The mini protein according to claim 1, characterized in that, The amino acid sequence of the mini protein is selected from SEQ ID No. 1-20, 50 and 54.

4. The mini protein according to claim 3, characterized in that, The amino acid sequence of the mini protein is SEQ ID No.

8.

5. A fusion protein, characterized in that, It is a fusion protein obtained by fusing the mini protein of any one of claims 1-4 with the serum albumin HSA binding domain ABD035.

6. The fusion protein according to claim 5, characterized in that, The amino acid sequence of the fusion protein is selected from SEQ ID No. 55 and SEQ ID No.

56.

7. A pharmaceutical composition, characterized in that, The pharmaceutical composition comprises a mini protein as described in any one of claims 1-4 or a fusion protein as described in claims 5 or 6, and pharmaceutically acceptable excipients.

8. Use of the mini protein of any one of claims 1-4, the fusion protein of claim 5 or 6, or the pharmaceutical composition of claim 7 in the preparation of an antagonist targeting IL-4Rα.

9. Use of the mini protein of any one of claims 1-4, the fusion protein of claim 5 or 6, or the pharmaceutical composition of claim 7 in the preparation of a medicament for treating allergic diseases.

Images