The LSPETase is an enzyme that effectively degrades PET. It was discovered by our research group led by PI through bioinformatics methods, which involved a series of PET hydrolases, including IsPETase. However, when we attempted to express LSPETase in Escherichia coli, we encountered the issue of inclusion bodies. To solve this problem, it is well known that adding signal peptides at the N- or C- terminus can change the protein folding environment and improve the expression of exogenous proteins. In our literature review, we found that Hogyun Seo et al. achieved extracellular expression of IsPETase in E. coli using a sec-dependent signal peptide[1], while Cui et al. achieved periplasmic expression using pelB[2]. However, there is a lack of research directly comparing the advantages and disadvantages of these two methods. Therefore, we established a model to simulate the extracellular and periplasmic environments, investigate the stability of LSPETase in these two different environments, assist wet experiments in selecting signal peptides, and attempt to elucidate the reasons for the improved stability of LSPETase at the molecular level
The extracellular environment, which mainly consists of bacterial culture media, differs significantly from the periplasmic space in terms of its oxidative nature. This oxidative nature can affect the formation of disulfide bonds in proteins, potentially influencing protein stability.
The extracellular environment, which mainly consists of bacterial culture media, differs significantly from the periplasmic space in terms of its oxidative nature. This oxidative nature can affect the formation of disulfide bonds in proteins, potentially influencing protein stability. To address this issue, we established a model to simulate the environments in the extracellular and periplasmic spaces and investigate the stability of LSPETase in these two different environments. The extracellular environment mainly consists of a bacterial culture medium, which differs from the periplasmic space in terms of its oxidizing nature. The oxidative nature of the periplasmic space may affect the formation of disulfide bonds in proteins, thereby influencing protein stability.
To simulate the denaturing factors experienced by LSPETase after expression, we used Pymol to remove ligands and water molecules from the PDB file and constructed a urea solution in Gromacs. The bond parameters of the two disulfide bonds in the top file were modified to simulate the effects of different oxidative environments. After energy minimization, temperature coupling, and pressure coupling, we performed a 50 ns simulation.
At a temperature of 300 K, we conducted MD simulations of LSPETase in 8 M urea solutions with Dij values of 0 and 25 using the single-precision version of Gromacs 5.1.4 and the GROMOS 54A7 force field. The structure and topology files of urea were generated using an automated topology builder and parameterized with the aforementioned force field. The protein molecule and SPC water molecules were placed in the box, and urea molecules were randomly added along with Cl- ions to balance the charge. Energy minimization was performed until the potential energy reached below 1000.0 kJ/mol/nm to ensure a reasonable starting structure for the entire system. Subsequently, the system was equilibrated in two steps: first using the V-rescale temperature coupling algorithm for 800 ps NVT equilibration, and then using the Parrinello-Rahman algorithm for 600 ps NPT equilibration. MD simulations were then run for 50 ns to obtain data.
To begin, we downloaded the topology and structure files for urea molecules from the Automated Topology Builder (ATB) website. We then generated the restraint information file for urea molecules[3]
Next, we constructed the box and inserted urea molecules into it. The amount of urea molecules to be inserted into the box was calculated using the equation mentioned above.
Afterward, we solvated the system and manually modified the top file to include the topology information of urea and match the number of molecules in the gro file.
Following energy minimization, temperature coupling, and pressure coupling, the resulting gro file served as the desired urea solution system file, which could be directly used as the input file for solvation.
1.1.3 Parameterisation of disulfide bonds
During the simulation, a Morse potential was used to describe the formation and breaking of disulfide bonds[4]
Dij represents the depth of the well, reflecting the oxidative capacity of the solvent environment. A lower Dij indicates a stronger reducing environment for protein folding, and vice versa. βij defines the steepness of the well and is fixed at 2.63nm-1. σ represents the equilibrium distance of the disulfide bond and is fixed at 0.38 nm
1.1.4 Feature Construction for MD
Our result analysis primarily relies on root mean square deviation (RMSD), root mean square (RMSF), native contacts, radius of gyration (Rg), hydrogen bonds (hbonds), and solvent accessible surface area (SASA). These parameters were chosen because they effectively distinguish native conformations from non-native conformations. RMSD was calculated by aligning the protein structures after NPT to a reference structure. Hydrogen bonds were considered formed if the donor-acceptor (D-A) distance was less than or equal to 3.5 Å and the A-D-H angle was less than or equal to 30°. Native contacts were defined as two atoms within 7 Å of each other in both the interested conformation and the native state. All parameters were calculated based on the entire protein.
1.2 Results of MD
Fig. 1 Rg
The radius of gyration is a dimensionless topological property of a molecule, the magnitude of which reflects the morphology of the molecule and the degree of interaction. In general, a larger radius of gyration indicates a tighter conformation of the protein molecule. Here it can be seen that the radius of gyration is higher under oxidising conditions.
Fig. 2 RMSD
RMSD measures the degree of conformational difference or trajectory stability. In the figure above, there is no significant difference in RMSD between the non-oxidizing and oxidizing conditions within the first 12000 ps, but the oxidizing condition exceeds the non-oxidizing condition afterwards[5].
、
Fig. 3 RMSF
RMSF calculates the fluctuation (magnitude of change) of each atom relative to its average position and serves as an indicator of the freedom of atomic motion and the flexibility of the molecular structure. In the amino acid residues 166-185, 197-215, and 257-271, the RMSF under non-oxidizing conditions is larger than that under oxidizing conditions.
Fig. 4 Amino acids 166-185 (A), 197-215 (B), and 257-271 (C) with their adjacent disulfide bonds
Amino acid residues 166-185, 197-215, and 257-271 are all located near disulfide bonds, indicating that the decrease in RMSF in these three peptide chains under non-oxidizing conditions is influenced by the oxidative environment affecting the disulfide bonds.
Fig. 5 SASA
SASA is the protein surface area that can be accessed by solvent. It can be observed that the SASA under denaturing conditions is slightly larger than that under oxidizing conditions.
Fig. 6 Number of hydrogen bonds within the protein molecule
During the simulation, the number of hydrogen bonds within the protein molecules was generally higher under oxidizing conditions. Hydrogen bonds are important factors that influence protein structure formation, and a higher number of hydrogen bonds may indicate a more stable protein state.
Fig. 7 Denaturing conditions (left) and oxidizing conditions (right) 2D density contour plots of RMSD and natural contacts
We constructed a two-dimensional contour plot using RMSD and native contacts as reaction coordinates. It can be seen that there is a significant difference in the distribution of conformations between these two conditions. Under fully denaturing conditions, the conformations are concentrated in multiple small regions, while under oxidizing conditions, the conformations are relatively evenly distributed with less deviation from the conformations. This is consistent with the characteristic of improved protein stability
In conclusion, we can infer that the oxidative environment enhances protein stability by affecting the disulfide bonds.
Visualization of the LSPETase simulation process under denaturing conditions
Simulating the LSPETase process under oxidation conditions