A method for constructing a model of aspart and glulisine based on molecular simulation
Patent Information
- Application Number
- CN202311555218.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-11-21
- Publication Date
- 2026-08-21
- Estimated Expiration
- 2043-11-21
AI Technical Summary
但是，面对新型蛋白分子时，在蛋白质数据库中有时找不到相应的pdb文件，也无法用GROMACS现有的力场去识别
[0004]本发明的目的是提供一种德谷与门冬胰岛素基于分子模拟的模型构建方法，打破了传统分子模拟模型构建的局限性，为德谷胰岛素与门冬胰岛素蛋白分子的模型构建提供了方法。
Smart Images
Figure CN117577221B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of molecular simulation technology, and in particular to a method for constructing a molecular simulation-based model of degludec and aspart insulin. Background Technology
[0002] Currently, with the diversification of medical technology, drug products can be personalized according to different individuals and needs. Insulin aspart (IASP) is a biosynthesized ultra-short-acting human insulin analog formed by replacing the proline at position 28 of the B chain of human insulin with aspartic acid. After subcutaneous injection, insulin aspart is rapidly absorbed, with an onset of action of only 5-15 minutes, maximum biological effect lasting 1-1.5 hours, and a duration of action of only 4-5 hours. Insulin degludec (IDEG), on the other hand, is an ultra-long-acting human insulin analog with a completely opposite effect to insulin aspart. It consists of a 22-carbon fatty acid diacid chain attached to the lysine residue at position 29 of the B chain of human insulin. Its onset of action is very slow, but its duration of action exceeds 42 hours. Insulin aspart and insulin degludec are protein drug products designed to meet different needs and belong to the category of novel proteins. To study the reasons for the differences in the mechanisms of action of novel proteins at the microscopic scale, molecular simulation is usually required for analysis.
[0003] In molecular simulations, traditional methods for constructing protein molecule models rely on two key elements: finding the appropriate PDB file and building the simulation force field. Traditionally, protein molecules can be modeled using PDB databases and other websites, with the force fields provided by GROMACS (AMBER, CHARMM, GROMOS, OPLS-AA) readily available. However, when dealing with novel protein molecules, the corresponding PDB file may not be found in protein databases, and existing force fields in GROMACS may not be applicable. Furthermore, the pH environments for drug preparation and release differ; for example, insulin degludec and insulin aspart are prepared in a strongly acidic environment, while their release conditions are in the human body, which is neutral. This makes it impossible to construct force fields for different pH values. These issues significantly limit the application of traditional protein molecule modeling to novel proteins. Summary of the Invention
[0004] The purpose of this invention is to provide a molecular simulation-based modeling method for degludec and aspart insulin, which breaks through the limitations of traditional molecular simulation modeling and provides a method for modeling degludec and aspart insulin protein molecules.
[0005] To achieve the above objectives, this invention provides a method for constructing a molecular simulation-based model of degludec and aspart insulin, comprising the following steps:
[0006] S1. By calculating the isoelectric point of the residues, determine the residue charges of insulin degludec and insulin aspart under different pH conditions. In the CHARMM36 force field file provided by GROMCAS, find the appropriate residue force field in the merged.rtp file, modify the residue force field name in the pdb file of insulin degludec and insulin aspart, and construct the residue force fields of insulin degludec and insulin aspart.
[0007] Among them, a suitable residue force field must satisfy: (1) the charge of the residue force field matches the actual residue charge under pH conditions; (2) the structure corresponding to the residue force field must match the actual residue structure;
[0008] S2. Add the atomic information of the fatty acid chain in insulin degludec to the force field of GROMACS, and modify the pdb files of insulin degludec and insulin aspart.
[0009] S3. Use GROMACS software to convert the pdb files of degludec insulin and aspart insulin into gro files, and generate topology files at the same time, indicating that the molecular simulation model has been successfully constructed.
[0010] Use VMD software to check for omissions and errors in the construction of molecular simulation models;
[0011] S4. Run the successfully constructed molecular simulation model in GROMACS using MD simulation to bring the energy and density of the insulin protein model to equilibrium. Extract the model conformation at this equilibrium state and use VMD to check the stability of the structure.
[0012] S5. Using umbrella sampling, molecular dynamics dragging was performed on the two insulin protein models to analyze and verify that the constructed insulin protein models conform to the real characteristics of insulin action.
[0013] Preferably, in step S2, the merged.rtp, merged.hdp, and residuetype.dat files in the CHARMM36 force field file that comes with GROMACS are modified to include the atomic information of the fatty acid diacid chain of insulin degludec into the above three files.
[0014] Preferably, the atomic information of the fatty acid chain includes: the charge information of the atoms, the bonding information between atoms, the matching information of hydrogen atoms with other atoms, and the type of fatty acid chain.
[0015] Preferably, in step S3, the pdb file is converted into a gro file using the pdb2gmx command in GROMACS. The full command text is as follows:
[0016] gmx pdb2gmx-fXXX.pdb-o XXX.gro-p topol.top-ss-merged all-ignh-ter-water spce;
[0017] In this configuration, -f represents the input pdb file name; -o represents the name of the generated gro file; -p represents the name of the generated topology file; -ss represents linking the disulfide bonds of insulin; -merged all represents merging the four insulin chains; -ter represents end-capping the ends of the insulin chains; and -water represents the model name for adding water molecules.
[0018] Therefore, this invention employs the aforementioned molecular simulation-based modeling method for degludec and aspart insulin. Simulation models of both insulins were constructed using GROMACS software. The models for both insulins under different pH conditions were then modified, and molecular dynamics (MD) simulations and umbrella sampling were performed. This method provides a modeling and force field correction approach for subsequent research on degludec and aspart insulins in the field of molecular simulation. Simultaneously, it overcomes the limitations of traditional molecular simulations, which can only construct models existing in protein databases and cannot directly provide force fields under different pH conditions.
[0019] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description
[0020] Figure 1 Diagrams showing models of insulin degludec and insulin aspart;
[0021] Figure 2 The graph shows the simulation results of insulin degludec in an acidic environment (pH=2.0);
[0022] Figure 3 The figure shows the simulation results of insulin degludec under neutral conditions (pH=7.4);
[0023] Figure 4 The graph shows the simulation results of insulin aspart in an acidic environment (pH=2.0);
[0024] Figure 5 The graph shows the simulation results of insulin aspart in a neutral environment (pH=7.4);
[0025] Figure 6 The graph shows the displacement of degludec and aspart insulin over time under traction. Detailed Implementation
[0026] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.
[0027] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.
[0028] Example 1
[0029] I. Construction of the force field in the degludec and aspart insulin model
[0030] (1) CHARMM (Chemistry at HARvard Macromolecular Mechanics) is a widely recognized and applied molecular dynamics simulation program used for the simulation of biological macromolecules, including energy minimization, molecular dynamics, and Monte Carlo simulations. It can provide users with empirical energy calculations for various small and large molecules (including proteins, nucleic acids, and sugars), including interaction energies and conformational energies. Local minimal structure files can be found in the protein database (RCSB ID: 4AKJ).
[0031] like Figure 1 The diagram shown is a model of insulin degludec and insulin aspart.
[0032] Because the 29th lysine residue of the B chain of insulin degludec is attached to a 22-carbon fatty acid diacid chain, the CHARMM36 force field cannot recognize it. Therefore, the primary task is to construct a force field that can recognize insulin degludec, and to modify the CHARMM36 force field accordingly.
[0033] (I) Open the merged.rtp file and add two sets of residue force fields to enable them to recognize lysine residues with carbon chains. Name them LYSD (lysine residue force field specifically for insulin degludec, suitable for acidic environments) and LYSDD (LYSD force field for deprotonated environments). LYSD is based on the LYS force field included in CHARMM36. Manually add 22 carbon atoms, 35 hydrogen atoms, and 6 oxygen atoms to the [atoms] column, along with the corresponding charge values. Since LYSD is a force field suitable for acidic environments with pH < 7.4, the addition is based on the requirement that the total charge sum must be a positive integer, such as +1, +2, etc. The charge value of a single atom is determined based on the charge values of atoms in similar amino acid structures. Then, manually add the bonding relationships between these carbon, hydrogen, and oxygen atoms to the [bonds] column, strictly adhering to the actual structure. LYSDD is a force field under deprotonation conditions, suitable for neutral environments with a pH of approximately 7.4. Unlike protonation conditions, the hydrogen atoms on the carboxyl groups are removed, so the number of atoms added in [atoms] is 22 carbon atoms, 33 hydrogen atoms, and 6 oxygen atoms, while the bonding atomic relationships are added in [bonds].
[0034] (II) Open the merged.hdb file and name it LYSD and LYSDD. Add the bonding relationships and number of hydrogen atoms. The basis for adding them is the carbon chain structure containing 22 carbon atoms, which includes the bonding relationships between hydrogen atoms, carbon atoms, and nitrogen atoms.
[0035] (III) Open the residuetypes.dat file and classify LYSD and LYSDD, assigning them to Protein. The atomic information of the additional fatty acid chains constructed in the Merged.rtp file is shown in Tables 1 and 2:
[0036] Table 1 [atoms]
[0037]
[0038]
[0039] Table 2 [bonds]
[0040]
[0041]
[0042]
[0043] (2) Next, use the pdb2gmx command in GROMACS to convert the pdb file of insulin degludec into a GRO structure file. During the conversion process, errors may occur indicating missing dihedral information. If this happens, find the corresponding atoms in the topology file based on the atomic numbers in the error message, determine the structural relationships between these atoms, and then formulate the dihedral information based on the atomic information of similar amino acid structures. Alternatively, you can submit it to the GCenFF website (CGenFF Home (umaryland.edu)) to generate reasonable dihedral information. After determining the dihedral information, add it to the ffbonded.itp file.
[0044] (3) If the pdb file of the degludec insulin protein molecule is successfully converted into a gro file, it indicates that the model has been initially established successfully.
[0045] (4) Insulin aspart (IASP) is formed by replacing the proline residue at position 28 of the B chain of human insulin with aspartic acid. Only residue substitution is performed; no additional atoms are added, making model construction simpler than for insulin degludec. Download the insulin aspart protein model (RCSB ID: 4GBC) from the protein database. Since insulin aspart does not have any additional atoms added, the file in the CHARMM36 force field does not need modification and can be directly converted using the pdb2gmx command in GROMACS.
[0046] II. Model Construction under Acidic (pH=2.0) and Neutral (pH=7.4) Environments
[0047] (1) In insulin degludec, some residues on the side chain have a +1 charge, such as LYS (lysine), ARG (arginine), and HIS (histidine); others have a -1 charge, such as ASP (aspartic acid) and GLU (glutamic acid). Based on the calculation of the isoelectric point of amino acids and the determination of the charge of each residue, we focused on observing these residues that already carry a charge in a neutral environment. It was found that in an acidic environment (pH=2), residues such as LYS, ARG, HIS, GLY, and PHE have a +1 charge, while Zn... 2+ The residues of LYS, ARG, and HIS have a +2 charge, while the GLU residues have a charge of 0. To construct a protein model under acidic conditions, a suitable residue force field needs to be selected. Therefore, reopen the merged.rtp file and search for the residue force fields of LYS, ARG, and HIS. The residue force field to be searched must meet two conditions:
[0048] (I) The charge of the residue force field must match the actual residue charge under acidic conditions.
[0049] (II) The structure corresponding to the force field of the residue must match the actual residue structure.
[0050] Based on these two conditions, suitable residue force fields were found: LYSD (LYS), ARG (ARG), HSP (HIS), and Zn2 (Zn 2+ ), GLUP(GLU), where the name inside the parentheses refers to the residue name, and the name before the parentheses refers to the force field name that the residue matches. LYSD only needs to be applied to the LYS residues on the B chain of amino acids. The D chain, since it does not have a 22-carbon fatty acid diacid chain, does not need to apply the LYSD force field; instead, the ordinary LYS force field is applied. All other residues must apply the found force field. It is important to note that although the GLY and PHE residues at the amino acid ends also have a charge of +1, their force fields do not need to be searched because they will be amino-terminated in the subsequent GROMACS instruction pdb2gmx.
[0051] (2) After finding a suitable residue force field, open the pdb file of insulin degludec (4AKJ.pdb), and replace all the force fields HIS and GLU in the file with HSP and GLUP, respectively. 2+ The force field was changed to Zn2; all atoms in the LYS chain and the additional fatty acid diacid were assembled together, and the force field was renamed LYSD.
[0052] (3) Use the pdb2gmx command in GROMACS to convert the pdb file of insulin degludec into a gro file. Successful conversion indicates that the force field model under acidic conditions has been successfully constructed. The force field construction in neutral and alkaline environments is the same as that in acidic environments. Perform the calculation and search according to the same steps, find the correct residue force field under the environment, change the force field name in the pdb file, and finally use GROMACS to successfully convert it into a gro file.
[0053] (4) When constructing models of insulin aspart under different pH conditions, the steps are basically similar to those for insulin degludec, except that the residue force field of ASP (aspartic acid) needs to be modified. Based on the calculation of the isoelectric point of the residues, the charge of ASP is 0 in an acidic environment. Therefore, it is necessary to find a suitable residue force field for ASP in the merged.rtp file under an acidic environment (pH=2), which is found to be the ASPP residue force field. Therefore, all the ASP force field names in the 4gbc.pdb file are changed to ASPP. The search and modification of other force fields are basically similar to those for insulin degludec. Finally, the 4gbc.pdb file is converted to a 4gbc.gro file using the pdb2gmx command in GROMCAS. Successful conversion indicates successful model construction.
[0054] III. Running the MD Simulation
[0055] Using GROMACS, we successfully constructed models of degludec / aspart insulin under acidic and neutral environments, and built a 6nm×6nm×6nm cubic box. Solvent water molecules (SPC / E) were added into the box, and Na+ and Cl- ions were added for neutralization. Subsequently, energy minimization and NVT and NPT equilibration were performed, and finally, the final product MD simulation was carried out.
[0056] Simulation results such as Figure 2 , Figure 3 , Figure 4 , Figure 5 As shown, the model remained structurally stable after the simulation, without exhibiting characteristics of chain detachment or structural damage (Note: Figure 2 , 3 In steps 4 and 5, due to periodic boundary conditions, some dimers appear to be separated, but are not actually separated.
[0057] IV. Umbrella Sampling
[0058] Insulin degludec and insulin aspart are ultra-long-acting and ultra-short-acting insulins, respectively, with significant differences in their release and action rates. Umbrella sampling molecular dynamics was performed on a constructed insulin degludec / aspart protein model. The AB chain of the insulin was fixed at pH 7.4, while the CD chain was dragged under traction force. The displacements of the two chains were compared under the same traction force and dragging time to reflect the difference in the ease of release between the two insulins.
[0059] like Figure 5 As shown, the drag displacement of insulin degludec was from 1.72nm to 1.98nm, a displacement of 0.26nm; while that of insulin aspart was from 1.66nm to 2.1nm, a displacement of 0.44nm, which is 1.7 times that of insulin degludec. This indicates that under the same conditions, the dimer of insulin aspart is more easily separated, so it is released from the dimer into monomers at a faster rate, which is consistent with the experimental data.
[0060] Therefore, this invention employs the aforementioned molecular simulation-based modeling method for degludec and aspart insulin. Simulation models of both insulins were constructed using GROMACS software. The models for both insulins under different pH conditions were then modified, and molecular dynamics (MD) simulations and umbrella sampling were performed. This method provides a modeling and force field correction approach for subsequent research on degludec and aspart insulins in the field of molecular simulation. Simultaneously, it overcomes the limitations of traditional molecular simulations, which can only construct models existing in protein databases and cannot directly provide force fields under different pH conditions.
[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.
Claims
1. A method for constructing a molecular simulation-based model of degludec and aspart insulin, characterized in that, Includes the following steps: S1. By calculating the isoelectric point of the residues, determine the residue charges of insulin degludec and insulin aspart under different pH conditions. In the CHARMM36 force field file included with GROMACS, find the appropriate residue force field in the merged.rtp file, modify the residue force field names in the pdb files of insulin degludec and insulin aspart, and construct the residue force fields of insulin degludec and insulin aspart. Among them, a suitable residue force field must satisfy: (1) the charge of the residue force field matches the actual residue charge under pH conditions; (2) the structure corresponding to the residue force field must match the actual residue structure; S2. Add the atomic information of the fatty acid chain in insulin degludec to the force field of GROMACS, and modify the pdb files of insulin degludec and insulin aspart. S3. Use GROMACS software to convert the pdb files of degludec insulin and aspart insulin into gro files, and generate topology files at the same time, indicating that the molecular simulation model has been successfully constructed. Use VMD software to check for omissions and errors in the construction of molecular simulation models; S4. Run the successfully constructed molecular simulation model in GROMACS using MD simulation to bring the energy and density of the insulin protein model to equilibrium. Extract the model conformation at this equilibrium state and use VMD to check the stability of the structure. S5. Using umbrella sampling, molecular dynamics dragging was performed on the two insulin protein models to analyze and verify that the constructed insulin protein models conform to the real function characteristics of insulin.
2. The method for constructing a molecular simulation-based model of degludec and aspart insulin according to claim 1, characterized in that, In step S2, the merged.rtp, merged.hdp, and residuetype.dat files in the CHARMM36 force field file that comes with GROMACS are modified to include the atomic information of the fatty acid diacid chain of insulin degludec into the above three files.
3. The method for constructing a molecular simulation-based model of degludec and aspart insulin according to claim 2, characterized in that, The atomic information of aliphatic diacid chains includes: the charge information of the atoms, the bonding information between atoms, the matching information of hydrogen atoms with other atoms, and the type of aliphatic diacid chain.