Application method of CCR5 binding protein with protein de novo design in HIV (Human Immunodeficiency Virus) inhibition

Abstract

The invention relates to a CCR5 binding protein with a protein de novo design and an application method of the CCR5 binding protein in HIV (Human Immunodeficiency Virus) inhibition. By adopting an artificial intelligence algorithm (such as RFdiversity, ProteinMPNN-FastRelax and AlphaFold2) and a molecular dynamics simulation technology, the invention provides an innovative de novo design method, and the protein molecule specifically bound with the CCR5 receptor can be accurately designed. The designed CCR5 binding proteins not only can effectively prevent the combination of HIV-1gp120 and CCR5 receptors, but also can inhibit the invasion of HIV viruses in the early stage of infection, and overcome the problems of drug resistance and side effects in the existing treatment methods. According to the design method, the production cost is reduced, the protein production efficiency is improved, a treatment strategy with a wide clinical application prospect is provided for HIV prevention and treatment, and remarkable economic value and market potential are achieved.

Description

Technical Field

[0001] Human immunodeficiency virus (HIV) infection depends on binding to host cell surface receptors. The CCR5 receptor is a key co-receptor for the entry of HIV virus into CD4+ T cells. Existing treatment methods such as CCR5 antagonists (e.g., maraviroc) and antibody drugs, although having achieved certain results in clinical practice, still have problems such as drug resistance, individual differences, and side effects. To solve these problems, the present invention proposes a new treatment strategy: applying de novo design of a CCR5-binding protein and precisely optimizing its structure to effectively inhibit the invasion of HIV virus. Background Art

[0002] Human immunodeficiency virus (HIV) is a virus transmitted through blood, sexual contact, mother-to-child and other routes, which can attack the host's CD4+ T cells, leading to the collapse of the immune system and thus causing acquired immunodeficiency syndrome (AIDS). The infection process of HIV depends on the binding of its surface glycoprotein gp120 to the host cell surface receptor CCR5. Especially in the R5 strain of HIV, the binding of gp120 to CCR5 plays a crucial role. As a key co-receptor for HIV invasion, CCR5 participates in the fusion of the virus and the host cell membrane, promoting the entry of the virus into CD4+ T cells and thus starting replication. Therefore, the CCR5 receptor has become an important target for HIV treatment. Especially in the research and development of antiviral drugs, finding a way to effectively inhibit the binding of the CCR5 receptor to HIV has become the focus of research.

[0003] Currently, CCR5 antagonists (such as maraviroc) and antibody drugs (such as PRO140) are one of the main strategies for treating HIV infection. Maraviroc inhibits the binding of HIV to CCR5 by blocking the CCR5 receptor, thereby preventing the virus from entering host cells. However, although maraviroc and other CCR5 antagonists have achieved certain results in clinical practice, these drugs still have some limitations. First, with the mutation of the HIV virus, some patients may develop drug resistance to these drugs, resulting in a gradual weakening of the treatment effect. Second, some CCR5 antagonists may cause adverse reactions, such as liver damage and immune system function inhibition, affecting the health and quality of life of patients. In addition, due to the complex structure of the CCR5 receptor, some virus variants may bypass the role of CCR5 through other routes, making the efficacy of existing treatment methods limited. Although gene therapy has also been proposed as an innovative treatment method, attempting to make host cells naturally immune to HIV invasion by editing the CCR5 gene (such as the CCR5-Δ32 mutation), this method faces high costs, technical complexity, ethical issues, and potential safety issues, and has not been widely applied clinically.

[0004] In recent years, with the rapid development of structural biology and artificial intelligence technologies, novel therapeutic strategies based on protein design have gradually emerged. Through means such as deep learning algorithms and molecular dynamics simulations, scientists can de novo design specific proteins that bind to the CCR5 receptor. These proteins can precisely prevent the binding of HIV to CCR5, with high affinity and specificity. Compared with traditional drugs, this method can not only overcome the problem of drug resistance but also has better stability and safety. Although this field is still in the research stage, its huge potential has gradually emerged and become an important research direction for anti-HIV treatment. Therefore, how to develop a new, efficient, and safe anti-HIV treatment strategy by using innovative protein design methods and optimizing CCR5-binding proteins with artificial intelligence algorithms has become a research hotspot. Summary of the Invention

[0005] The present invention provides a method for de novo designing CCR5-binding proteins based on artificial intelligence algorithms and molecular dynamics simulations. This method can effectively inhibit the entry of HIV virus into host cells through the CCR5 receptor, thereby effectively reducing the risk of HIV virus infection. The design method of the present invention adopts advanced deep learning algorithms (such as RFdiffusion, ProteinMPNN-FastRelax, AlphaFold2) and molecular dynamics simulation techniques. By precisely optimizing the key binding sites at the binding interface between the CCR5 receptor and gp120, it overcomes the problems of drug resistance, unstable efficacy, and large side effects in existing anti-HIV treatment methods.

[0006] The core technical solutions of the present invention include:

[0007] 1. Protein de novo design method: Use artificial intelligence algorithms such as RFdiffusion, ProteinMPNN-FastRelax, and AlphaFold2 for de novo design of proteins. Through deep learning algorithms, design proteins that can bind to the CCR5 receptor from scratch, using the existing gp120, CCR5, and CD4 complex protein (PDB ID: 6met) as a design template to ensure that the designed proteins have high affinity and specificity.

[0008] 2. Optimize the binding sites: During the design process, focus on optimizing the key sites where the CCR5 receptor binds to gp120, especially regions such as the V3 loop of gp120, the 310-Gly-Pro-Gly-Arg(Gln)-313 motif, and the N-terminal residues of CCR5. These binding sites play an important role in the binding of CCR5 and gp120. By optimizing these sites, the designed proteins can efficiently block the binding of HIV to the CCR5 receptor.

[0009] 3. Molecular Dynamics Simulation and Screening: The designed protein is structurally optimized through molecular dynamics simulation to further improve its stability, affinity, and binding ability to the CCR5 receptor. After screening, proteins with the best affinity and low pAE_interaction values (<10) are selected as the final candidate molecules.

[0010] 4. Protein Expression and Purification: The designed protein is expressed through yeast surface display technology, and a highly purified CCR5-binding protein is obtained through an optimized purification process. This method can not only efficiently express the protein but also produce it on a large scale to meet the requirements of clinical applications.

[0011] 5. HIV Inhibition Application: The designed CCR5-binding protein is applied to the treatment of HIV, which can effectively prevent the HIV virus from entering host cells through CCR5 and prevent CD4+ T cells from being infected by the virus. The affinity of the designed protein and its effect on inhibiting HIV are verified through technologies such as Biolayer Interferometry (BLI) and flow cytometry.

[0012] The beneficial effects of the present invention are as follows:

[0013] (1) Overcoming the drug resistance problem of existing anti-HIV treatments: Through the de novo designed CCR5-binding protein, the present invention can effectively prevent the binding of HIV-1 gp120 to the CCR5 receptor, avoiding the drug resistance problem that occurs during the long-term use of traditional drugs (such as CCR5 antagonists). Since the protein of the present invention is designed based on artificial intelligence algorithms and molecular dynamics simulation, it can precisely optimize the binding to CCR5 and reduce the impact of virus mutations on the treatment effect.

[0014] (2) Improving the specificity and affinity of treatment: The present invention precisely optimizes the key binding sites of the CCR5 receptor, such as the V3 loop, the 310-Gly-Pro-Gly-Arg(Gln)-313 motif, and the N-terminal residues of CCR5. Compared with traditional drugs, Designed The binding ability of the protein is increased by 500 times , it can efficiently bind to the CCR5 receptor and prevent the HIV virus from invading host cells. This specific design greatly improves the therapeutic effect of the protein and reduces non-specific binding and side effects.

[0015] (3) High stability and significant biological effects: The CCR5-binding protein designed in the present invention is structurally optimized through molecular dynamics simulation to ensure its stability and good performance in vivo and in vitro. Through experimental verification, the designed protein has a significant effect on inhibiting the invasion of the HIV virus, can effectively prevent the virus from infecting CD4+ T cells, and reduce the replication ability of HIV.

[0016] (4) Low-cost production process: The present invention uses yeast surface display technology for protein expression, and by optimizing protein expression, purification, and production processes, it reduces the production cost of proteins and improves production efficiency. Compared with traditional antibody drugs or gene therapies, the present invention provides a more economical and large-scale producible anti-HIV treatment solution.

[0017] (5) Wide clinical application prospects: The CCR5-binding protein of the present invention can not only be applied to the treatment of HIV, but also has the potential to be applied to the treatment of other viral infections or immune system diseases. In addition, due to its low production cost and stable efficacy, it can become an anti-HIV drug widely used globally, especially providing a new solution for HIV treatment in developing countries and regions with limited resources.

[0018] (6) Provide a new approach for HIV cure: The CCR5-binding protein provided by the present invention can not only effectively control HIV infection, but also provides a new technical path for future HIV cure research. By blocking the binding of the virus to CCR5, effective virus clearance and long-term inhibition can be achieved, providing important support for the implementation of HIV cure strategies. Brief Description of the Drawings

[0019] Figure 1 For the two-dimensional structure, three-dimensional structure of CCR5, and the binding site of HIV-1 gp120 and CCR5

[0020] [Note: A is the two-dimensional structure of CCR5 with seven transmembrane segments. The left red dashed box is the extracellular N-terminal fragment, and the right dashed box is the second extracellular loop (ECL2), which is the strong binding site of HIV-1 gp120 and the CCR5 receptor; B is the three-dimensional structure of the CCR5 receptor (cited); C is the electron microscopy structure of gp120 (RCSB Protein Data Bank code (PDB ID): 5VN328), CCR5 (PDB ID: 5UIW10), and the four-domain CD4 (PDB ID: 1WIO27) (cited)];

[0021] Figure 2 For de novo design of 5000 binding proteins using the RFdiffusion diffusion model

[0022] (Note: A shows that the design of 5000 binding proteins has been completed; A is the structure of the designed protein)

[0023] Figure 3 For optimization and filtration by ProteinMPNN-FastRelax and AlphaFold2

[0024] (Note: A has completed the optimization and sequence generation of 5000 proteins ProteinMPNN; B is the design of sequences and scoring for binding proteins)

[0025] Figure 4 Flowchart for the preparation of the CCR5-binding protein by de novo design of proteins of the present invention Detailed implementation manner

[0026] This example details the process of de novo designing the CCR5-binding protein, which mainly includes steps such as protein design, optimization, screening, and molecular dynamics simulation. By using artificial intelligence algorithms and molecular dynamics simulation techniques, we precisely optimized the structure of the CCR5-binding protein, aiming to improve its affinity, specificity, and stability with the CCR5 receptor.

[0027] 1. Protein design and selection of template

[0028] First, select the gp120, CCR5, and CD4 complex protein (PDB ID: 6met) as the basic template for design. This complex provides detailed structural information on the binding of gp120 to the CCR5 receptor, which can provide us with precise data on the binding interface between CCR5 and gp120. By utilizing this structure, we can identify key binding sites and optimize them.

[0029] During the design process, we selected the following key binding sites for optimization: V3 loop: The V3 loop of gp120 is the main region for the binding of HIV to CCR5, containing a conserved 310-Gly-Pro-Gly-Arg(Gln)-313 motif. This motif penetrates about one-third of the thickness of the lipid bilayer and enters the CRS2 pocket of CCR5, and the Pro311 residue reaches the deepest part. This motif plays a crucial role in the binding of the virus to the receptor. N-terminal residues of CCR5: The N-terminal region of CCR5 and the bridging fragment of gp120 together form the second main binding interface for virus invasion, especially helices I and VII, which are stably connected by a disulfide bond (Cys20–Cys269), and the N-terminal fragment (residues 1-19) connected to helix I also plays a key role in the binding process ( Figure 1 A).

[0030] 2. Use RFdiffusion for protein design

[0031] We used RFdiffusion for de novo design. RFdiffusion is a deep learning-based protein structure design algorithm that can generate new protein structures based on a given structural template (in this case, PDB ID: 6met) and further improve their binding ability to the CCR5 receptor through optimization design. In this step, we selected the following hotspots for optimization:

[0032] B1-19 (N-terminal fragment of CCR5)

[0033] B167-198 (other important regions where gp120 contacts the CCR5 receptor)

[0034] B10, B14, B86, B108, B172 (other important regions where gp120 contacts the CCR5 receptor)

[0035] B251-283 (binding site of the CCR5 receptor)

[0036] RFdiffusion generated 5000 new protein structures that contain the optimized binding sites and hotspots. Each newly designed protein followed the principles of de novo design and was optimized according to the given target structure to ensure its high affinity and high specificity for the CCR5 receptor ( Figure 1 A,B).

[0037] 3. Molecular Dynamics Simulation and Protein Optimization

[0038] The 5000 newly generated protein structures were further optimized through molecular dynamics simulation (MD simulation). MD simulation helps to evaluate the stability of the protein in different environments and its binding ability to the CCR5 receptor. Through MD simulation, we were able to: determine the conformational changes of the protein when binding to the CCR5 receptor; evaluate the binding strength between the designed protein and the CCR5 receptor; optimize the spatial structure of the protein to ensure that it can bind stably and interact effectively with the CCR5 receptor. After optimization through MD simulation, we selected the protein structures with the best binding affinity and stability and further screened them ( Figure 2 A,B).

[0039] 4. ProteinMPNN-FastRelax and AlphaFold2 Screening

[0040] After MD simulation optimization, we further screened proteins using ProteinMPNN-FastRelax. ProteinMPNN-FastRelax optimizes the amino acid sequence by generating multiple sequences for each backbone, enabling the protein to enhance its binding ability to the CCR5 receptor while maintaining a stable structure. We used AlphaFold2 for initial guess and target templating screening. AlphaFold2 is a deep learning model that can predict the three-dimensional structure of a protein based on its amino acid sequence and select the most suitable structure in templating screening. After screening, CCR5-binding proteins with pAE_interaction < 10 were finally obtained, indicating that the binding affinity of these proteins to CCR5 reached the expected target( Figure 3 A,B).

Claims

1. A method for preparing a de novo designed CCR5-binding protein, characterized in that, The method includes the following steps: (1) Use artificial intelligence algorithms for de novo protein design. The artificial intelligence algorithms are selected from deep learning algorithms such as RFdiffusion, ProteinMPNN-FastRelax, AlphaFold2, etc.; Use the gp120, CCR5, and CD4 complex protein (PDB ID: 6met) as the design template to optimize the key binding sites of the CCR5 receptor. The binding sites include the V3 loop of gp120, the 310-Gly-Pro-Gly-Arg(Gln)-313 motif, and the N-terminal residues of CCR5, as well as the hot spot regions B1-19, B167-198, B10, B14, B86, B108, B172, B251-283; Obtain 5000 new protein structures through molecular dynamics simulation and optimize the designed CCR5-binding protein; (2) Use ProteinMPNN-FastRelax to generate two sequences for the backbone of the protein, and use AlphaFold2 for initial guess and target templating screening. After screening, obtain CCR5-binding proteins with pAE_interaction < 10; (3) Perform protein expression and purification through yeast surface display technology to obtain CCR5-binding proteins with relatively high purity.

2. The method according to claim 1, wherein the optimization step includes optimizing the key binding sites of the CCR5 receptor. The binding sites include the V3 loop, the 310-Gly-Pro-Gly-Arg(Gln)-313 motif, and the N-terminal residues of CCR5, as well as the hot spot regions B1-19, B167-198, B10, B14, B86, B108, B172, B251-283.

3. The CCR5-binding protein according to claim 1 or 2, characterized in that, The protein can effectively bind to the CCR5 receptor, prevent the binding of HIV-1 gp120 to CCR5, and thus inhibit the invasion of HIV virus.

4. A method for applying the CCR5-binding protein according to claim 1 or 2, characterized in that, The method includes: (1) Apply the CCR5-binding protein to the treatment of HIV to prevent the HIV virus from entering host cells through CCR5; (2) Determine the affinity between the CCR5-binding protein and the CCR5 receptor through Biolayer Interferometry (BLI); (3) Detect the inhibitory effect of the CCR5-binding protein on the binding of gp120 of HIV-1 to CCR5 by flow cytometry.

5. The application method according to claim 4, wherein the HIV-1 is R5-type HIV virus, and the CCR5-binding protein can effectively inhibit the invasion of the virus and prevent CD4+ T cell infection.

6. A method for producing a CCR5-binding protein according to any one of claims 1 to 6, characterized in that, Use a low-cost expression system for large-scale production, and reduce production costs and improve production efficiency by optimizing protein expression, purification, and production processes.

7. A pharmaceutical composition for treating HIV, characterized in that, Contain the CCR5-binding protein according to claim 3 or 5 as an active ingredient for inhibiting HIV infection.

8. The pharmaceutical composition according to claim 7, wherein the pharmaceutical composition can be administered by injection, oral administration, or topical application.

Images