Molecular Dynamics Study of Citrullinated Proteins Associated with the Development of Rheumatoid Arthritis
Description
Results of the molecular dynamics experiment with the structural blocks of target proteins, the modification of which is associated with rheumatoid arthritis development
Files
Steps to reproduce
We selected proteins with citrullination localized in a compact structural element that could be stable independently of the entire macromolecule. Protein structures were obtained from the RCSB PDB (PDB ID–1BIK, 1N5U, 3E1I, 6ANU) and AlphaFold Protein Structure Database (AF-H9GXR8-F1). Super secondary elements were cut using PyMol software (Schrödinger, Inc., city, NY, USA) for the following steps of the present study. The molecular dynamics simulations were performed using the GROMACS 2020.4 software and the modified GROMOS96 54A7 force field in an explicit solvent under periodic boundary conditions. Each starting structure was centered in a cubic box of sufficient size environment so that the minimum distance to period images was at least 1.0 nm. A sim-ple point charge water model was employed in the simulations. The systems were neutralized with sodium (Na+) and chlorine (Cl−) ions. Protein and non-protein atoms were coupled to their temperature baths set at 311 K using the V-rescale algorithm. The pressure was maintained isotropically at 1 bar using a Berendsen barostat. A time step of 2 fs was employed. Each system underwent energy minimization using the steepest descent algorithm (1000 steps) followed by gradual heating from 5 to 311 K dur-ing a 200 ps molecular dynamics run with fixed heavy atoms of structural elements. Initial atom velocities were taken from a Maxwellian distribution at 311 K, and the bond lengths were con-strained using LINCS. A 1.4 nm cut-off was used for Lennard–Jones interactions, and dispersion corrections for energy and pressure were applied. Electrostatics interactions were calculated using the particle-mesh Ewald method with a 0.12 nm grid-spacing and a 1.4 nm real-space cut-off. Each trajectory was run for 20 ns in three repetitions, with a total simulation time of 0.6 μs. Each simulation system was run for 20 ns in three repeti-tions, with a total simulation time of 0.6 μs. Molecular dynamics trajectories were analyzed using the standard GROMACS utilities. The root-mean-square deviation (RMSD), solvent-accessible surface area (SASA), and pairwise distances were calculated using GROMACS built-in tools. Cluster analysis was performed for all structural elements using the ‘gmx’ cluster module, where the backbone atoms of all residues were superimposed. The Gromos clustering method was applied, and the largest clusters were extracted. Major conformations were analyzed and compared using the 2StrucCompare web server in the STRIDE al-gorithm.