CHAPTER THREE
3. WILD TYPE - MOLECULAR DYNAMICS
3.3 METHODOLOGY
files. The lack of parameters for cofactors also challenge the use of force-fields in many simulations.
However, webservers such as ATB (Malde et al., 2011), ProDRG (van Aalten et al., 1996) and ACPYPE (Sousa da Silva and Vranken, 2012) have been developed to bridge this gap.
H57, H71 and D49 including the Iron) forcefield parameters were provided by the previous group (Sheik Amamuddy et al., 2020). The AMBER FF was used to perform all atoms MD simulation (AMBER SQM V19) (Walkeret al., 2008). The importance of all atom MD simulations is to study and explore conformational flexibility and stability of protein and/or ligand systems (Musyokaet al, 2016).
3.3.2 MD simulation runs
3.3.2.1 Protein and ligand preparation
The WT PZase protein was protonated on H++ webserver (http://biophysics.cs.vt.edu/) at pH 7. The other parameters were left at default of 0.15 salinity, 10 internal dielectric, 80 external dielectric and no preparation of explicit solvent box topology/coordinate files (AMBER). The obtained topology and coordinate files were concatenated to a pdb file using theambpdbcommand.
An ad hoc Python script was used to reduce the ligand, delete extra hydrogens, rename the HIS and ASP atoms to HID/HIE and AP1 respectively, correctly number all atoms, merge the protein, iron and ligand coordinates files and execute tLeap and AMBER ACPYPE (antechamber Python parser interface). tLeap and ACPYPE were used to infer parameters and import AMBER files to generate GROMACS compatibles. The tLeap step solvated the system using TIP3P mode (Mahoney and Jorgensen, 2000).
3.3.2.2 Energy minimization
Energy minimization was done using the GROMACS (v2018) gmx mdrun to avoid steric clashes that might have formed due to inappropriate geometry and/or addition of ions and water molecules.
This step was done so as to relax the system using the steepest descents method with a force tolerance of 1000 kJ/mol/nm capped to an upper limit of 50 000 steps.
3.3.2.3 Equilibration
After minimization, the systems were equilibrated to mimic the biological environmental conditions necessary for the functioning of the protein. Equilibration of temperature and pressure were done on CHPC cluster. The temperature of the system was equilibrated using a constant number of particles, volume and temperature (NVT) ensemble over a period of 100 ps at 300 K while pressure was equilibrated at 50 000 steps for 100 ps at 1 atm using the NPT ensemble.
3.3.2.4 MD simulation
CHPC cluster was used for the dynamic simulations of the protein complexes using GROMACS (gromacs/v2016). The initial 93 simulations were run for 20 ns at 2 fs timestep. From the 93 simulation systems, compounds that portrayed unimodal conformations (47 compounds) in the last 10 ns of the 20 ns simulation period were extended to 150 ns using the same timestep.
3.3.3 Post MD trajectory analysis
After completion of 150 ns MD simulations, the whole system was removed from the periodic boundary conditions and centered using the MD trajectory analysis (trjconv) tool. Analysis of the
trajectories were computed using Radius of gyration (Rg), hydrogen bond,RMSD and RMSF. The dynamics of the systems throughout the simulation period were visualized in Visual Molecular Dynamics (VMD) (Humphrey et al., 1996). The ad hoc analysis Python and R scripts used were provided by the previous group (Sheik Amamuddy et al., 2020), the codes were edited to suit the current data.
3.3.3.1 RMSD
RMSD measures the average distance between atoms of superimposed proteins. It is often used in globular protein studies measuring similarities of alpha carbon atomic co-ordinates and molecular configuration of ligands when bound to macromolecules (Zhaoet al.,2015). To determine how each ligand was behaving throughout the simulation, initial analysis of the ligand RMSD was done. The ligands that portrayed stable and unimodal conformations were identified and reviewed in further analysis. Alpha carbon RMSD was also computed. A Python script was used to create compatible R data files which were analysed in R studio to generate a data frame for use in Jupyter Lab. The RMSD data was then plotted as violin plots using Jupyter Lab.
3.3.3.2 RMSF
RMSF measures local chain flexibility by calculating the deviation of protein residues from the averaged position of the particle over time (Zhao et al., 2015). To monitor the protein residues motions in each system, RMSF was computed based on alpha carbon atoms using the GROMACS rmscommand. A Python script was used to create compatible R data files which were analysed in R studio to generate a data frame for use in Jupyter Lab. The RMSF data was then plotted as a heat map in Jupyter Lab.
3.3.3.3 Radius of gyration
Radius of Gyration measures the protein compactness by calculating the distance between protein centre of mass and its terminals. A stably folded structure roughly maintains a steady Rg value. The whole protein compactness and active site compactness were computed. Active site Rg was computed by selecting residues interacting within 8 Å of the PZA ligand. The results were represented as violin plots plotted in Jupyter Lab after the use of a Python script and R compatible files in R studio.
3.3.3.4 Hydrogen bonding profiling
Since a hydrogen bond is one of the main stabilizing forces in molecular structures, the number of present hydrogen bonds were computed throughout the simulation period using the GROMACS gmx hbondcommand. Python scripts and R studio were used to extract and analyse the data while Jupyter Lab was used to construct thehbondplot. To determine the precise residues that formed the hbonds, thecpptraj4command was used.
3.3.3.5 VMD visualization
Protein dynamics results were visualized on VMD, a tool used to animate and analyze trajectories of MD simulations.