3.1. Evaluating deformation of spike protein structure by MOFs
For evaluation, the snapshots of each simulation's initial and final stages are provided in Fig. 2A. As can be seen, MOFs have all been attracted to the surface of S protein. At first glance, it is clear that ZIF covers the most surface area of the S protein, whereas UIO and IRMOF have lower attached surfaces to the S protein. It implies that ZIF interacts with the protein more successfully. In the following sections, all simulations' quantitative evaluations are pursued to understand the interactions better.
Figure 2B represents the MOFs density and water density in the simulation boxes. The horizontal arrangement of the MOF on the S protein is confirmed by the more even density diagram of ZIF across the box. On the other hand, sharp peaks of UIO and IRMOF across the axes confirm the vertical positioning of the structures with a lower interface with S proteins.
Following the interactions with nanomaterials, the structure of S protein can change. To examine the deformation, the distribution of the secondary structures for S proteins after interaction with MOFs has been evaluated (Fig. 3A). Increasing of β-sheets and α-helices intensity and reduction of the coils, bends, and turns indicate the stable structure with more likely interaction with ACE2. As shown in the diagram, pristine S protein with no interaction with MOFs has the most β-sheets in its structure, which can be interpreted as its stable structure and, as a result, its higher capability in interacting with ACE2.S protein after contact with ZIF exhibits the lowest β-sheets (21%) without α-helix in the structure that displays ZIF’s inhibitory effect against S protein function. Moreover, loosely structured proteins (e.g., coil and turn) hit the maximum amount in the secondary structure of S protein after contact with ZIF.
To determine the effect of three various MOFs on S protein-ACE2 complex formation, we evaluated van der Waals (vdW) and electrostatic interactions between S protein and the structures. Since the vdW and electrostatic energies between the molecules are important during structural deformations, these interactions are studied. Fig. 3B presents the average energy of vdW and electrostatic interactions between S protein and each MOF. The vdW forces outnumber the electrostatic forces, which is more pronounced in the total energy. ZIF has the highest total binding energy of any spike protein or MOF. The stronger interactions between S protein and ZIF resulted in more deformation in the protein structure by recent MOF. The vdW energy is an accurate indicator of hydrophobic forces. The amount of this force is related to the atomic radius of the MOF. As a result, vdW energy is usually greater at larger atomic radii. This is supported by ZIF's higher negative energy and larger radius. S protein structural changes, on the other hand, are accompanied by an increase in stabilizing structures and a decrease in destabilizing structures as a result of its interaction with ZIF. Low vdW energy and appropriate structural changes are two essential criteria for ZIF to better inhibit this protein.
H-bonds are among the most important intermolecular forces. Increased hydrogen interactions between S protein and MOFs can also alter the structure of S protein and reduce its interactions with ACE2. The average of H-bonds formed between S protein and MOFs is shown in Fig. 3C. The highest and lowest H-bonds with S protein were found in the ZIF and IRMOF structures, respectively. Since these interactions can deform the S protein structure, ZIF is the most effective structure on the S protein secondary structure's instability compared to other investigated MOFs. In this regard, increasing the hydrogen interactions between S protein and MOFs reduces this protein's interaction with aqueous media and reduces its contact area with aqueous media. The average contact area between S protein and MOFs is depicted in Fig. 3C. It reveals that in the presence of ZIF and IRMOF, S Protein had the lowest and highest contact area with aqueous media, respectively. This is due to the high level of hydrogen bonds between ZIF and S protein, which deforms SARS-critical CoV-2's protein structure. The lower the solvent available surface area (SASA), the less nanoparticle solvent available. On the other hand, the greater the number of H-bonds, the stronger the bond between the nanoparticle and the protein. As a result, ZIF has the best interference. Baweja et al. 45 investigated the interaction of a specific protein folding with graphene-based nanoparticles. The result shows an increase in the number of H-bonds and a decrease in the SASA of graphene oxide interaction with protein.
3.2. Evaluation of the effect of S protein deformation on its interaction with ACE2
In the previous section, the effect of MOFs on the deformation of the S protein structure was investigated. The current section discusses how S protein deformation affects its interaction with ACE2. Fig. 4A reveals the structure of S protein (blue) and ACE2 (green) as well as their interaction energies in the presence of MOFs after docking simulations. The use of deformed S protein structures reduced the energy of the interaction with ACE2 and increased the distance between the S protein and ACE2. Among the considered MOFs, the interaction between ACE2 and ZIF-deformed S protein had the lowest docking energy. Therefore, docking results confirm the previous findings and identify ZIF as the best structure for inducing the S protein structure's deformation. In this regard, the difference in initial and final entropy caused by the interaction of S protein and ACE2 was also investigated and shown in Fig. 4B. The greater the entropy difference, the greater the negative Gibbs free energy, resulting in a more stable S protein interaction with ACE2.Although deformation of S protein by MOFs reduced the entropy difference, the interaction between ACE2 and S protein deformed by ZIF had the lowest entropy difference. This indicates that the S protein-ACE2 complex is unstable due to the deformation of the S protein structure. Entropy analysis, like docking results, shows ZIF as the best structure to deform S protein.
On the other hand, the interaction between S protein and ACE2 causes more compactness of the protein structure. The degree of S protein compaction is revealed by analyzing the gyration radius. The greater the S protein compactness, the smaller the gyration radius. So, in this study, we evaluated the difference in the radius of gyration at the initiation and the end of the simulation as a comparison index of S protein compaction (Fig. 4B). A negative difference in the radius of gyration indicates decreased compactness of the deformed S protein after the interaction with ACE2. The littlest interaction between ACE2 and the S protein deformed by ZIF is observed. This deformed S protein in this simulation had the smallest difference in the radius of gyration. Mousavi et al. 46 investigated the conformational behaviors of chitosan nanoparticles on donepezil and rivastigmine drugs. By varying the ions, the Rg of drugs and polymers was altered. The Rg decreases, and the stability increases as the drug loading increases. In this regard, our findings show that ZIF has the smallest radius and thus the most stable interference. This nanoparticle has the lowest energy and the most stable state in terms of energy. As previously stated, H-bonds are one of the most powerful intermolecular interactions and significantly impact the intermolecular bonds between S protein and ACE2. As a result, studying the H-bonds between S protein and ACE2 is a good indicator of the effects of S protein deformation on its interaction with ACE2. The average of the H-bonds formed between S protein and ACE2 is shown in Fig. 4C. According to the findings, deformation of S protein by MOFs reduced H-bonds, indicating the effectiveness of S protein deformation in reducing the interaction with ACE2. Because ZIF was more effective at reducing hydrogen interactions, it is the best structure for deforming S protein.
3.3. Main protease structural variation after interaction with 3D MOFs
The importance of SARS-CoV-2 Mpro in the virus replication cycle was explained. The effects of 3D structures, including ZIF, IRMOF, and HKUST, on various secondary structures of the enzyme, were investigated, and snapshots from the last stage of simulations are provided in Fig. 5A. The amount of each secondary structure of the enzyme, including β-sheets, helices, β-bridges, turns, bends, and coils, changed after interaction with all investigated MOFs. The distribution of the secondary structures of Mpro (Figure 5B) demonstrates that the percentage of the coil, turn, and bend structures of the enzyme increased after interaction with the mentioned MOFs. HKUST had the greatest increase in coil, turn, and bend structures. Thus, all 3D MOFs weakened the enzyme structure stability compared to the control group (pure enzyme), while HKUST induced instability more than other 3D materials. As is well-known, surface engineering and modification can improve the properties and performance of nanomaterials. In this regard, we investigated the effect of the hydroxyl group on the HKUST as the MOF with the best performance in the considered group. As expected, functionalized HKUST (HKUST-OH) yielded the highest degree of instability, even more than pristine HKUST. As is obvious, the structures that cause the most changes in the enzyme structure, from most to least, are as follows: HKUST-OH, HKUST, IRMOF, and ZIF. In a similar study, Jin et al. 47 investigated the effect of graphene oxide nanosheets on the secondary structure of β-amyloid using DPPS analysis. Exposure to graphene oxide nanosheets increased the percentage of coil structures and decreased the percentage of β-sheets of β-amyloid. Simulation results show that nanomaterials destabilize the protein structure, which is consistent with their research.
To gain deep insight into the impact of nanomaterials on the SARS-CoV-2 Mpro using g_mmpbsa software 48, the interactions are analyzed from an energetic point of view, including vdW and electrostatics as well as total energy (Fig. 5C). Negative energy values indicate stable interaction between the corresponding nanomaterial and the unaffected enzyme (as the control group). Surface modification of HKUST by adding hydroxyl groups led to boosted vdW and electrostatic interactions. It can be attributed to the enlarged HKUST structure due to the presence of functional groups that consequently strengthen vdW attractions. Furthermore, because of the presence of negative -OH groups, the electrostatic energy was amplified, causing the protein to be more strongly adsorbed toward HKUST-OH. On the other hand, the vdW interaction of Mpro with IRMOF is significantly stronger than with other MOFs. It can be explained by the presence of iron in this framework, which increases the structure's vdW radius and thus amplifies this attraction. However, in the case of HKUST (-OH), the electrostatic interactions also add up to the total interactions and cause stronger attractions of HKUST(-OH) with the protein. Altogether, surface modification of the HKUST modifies both vdW and electrostatic adsorptions, i.e., surface engineering plays a critical role in the design of nanomaterials against COVID-19.
SASA (during the simulation and average values) for the SARS-CoV-2 Mpro with or without (control group) nanomaterials is shown in Fig. 6A. As can be seen, the SASA amount is lowest for Mpro in the presence of HKUST-OH, indicating a shorter distance between the nanomaterial and the enzyme. In other words, Mpro is mostly in contact with HKUST-OH rather than being exposed to solvent. Therefore, the interaction between HKUST-OH and the enzyme is stronger among its peers. To compare the relative exposure of Mpro in the presence of nanomaterials, the average SASA for all cases is provided in Fig. 6A-ii. Apparently, the use of pristine HKUST and even surface engineered HKUST-OH reduce SASA of SARS-CoV-2 Mpro i.e., accessible surface area of the enzyme and its functionality is reduced.
As variation in Rg decreases, the systems become denser and more petite. The greater the difference, the smaller and denser in similar systems (Fig. 6B-i). As it can be seen in Figure 6B-ii, the difference between the final and initial Rg is lowest for HKUST-OH, indicating the best interference for these MOFs. Rg analysis results confirm the previous SASA results. Chen et al. 49 used molecular dynamics simulation to demonstrate that the addition of graphene oxide nanosheets reduces the Rg of beta-amyloid. Our findings confirm the findings from SASA analysis on the reduced surface area and functionality of SARS-CoV-2 Mpro, which are consistent with their findings. The radial distribution functions (RDF) parameter that can be obtained from different methods is used to investigate the molecular aggregation at a specific simulation box location. Compared to molecular aggregation between several systems, the higher the maximization of this factor, the greater the system's molecular aggregation. According to Fig. 4F, the highest values of RDF are detected with HKUST-OH, HKUST, IRMOF, and ZIF, respectively. Therefore, the complex of HKUST-OH and Mpro has the best molecular aggregation and accumulation. In another study, Kamel et al. 50 investigated the effect of different amino acid adsorption on the functional and non-functional nanoparticles. In contrast to our present work, they obtained less RDF for interference between amino acids and functional nanoparticles than non-functional nanoparticles. These differences in graphs can be related to the intrinsic properties of materials and groups.
The impact of nanomaterials on the SARS-CoV-2 Mpro is evaluated through entropy calculations. In this regard, the entropy of each simulation is computed at different stages of interaction (Fig. 6C-i). For all cases, the entropy of the system increases by progress in time. However, for the simulations of SARS-CoV-2 Mpro in HKUST’s presence, the increase is more accentuated. Entropy increases with surface modification of HKUST with hydroxyl groups, indicating that the enzyme's functionality has been lost due to interactions with MOF molecules.
The presence of hydrogen atoms attached to electronegative atoms (such as fluorine, oxygen, and nitrogen) is important to form H-bonds. The amount of H-bonds between the enzyme and the nanomaterials indicates the bonding as well as the strength of the interference between them. This analysis is a vital and exciting criterion for predicting the interaction between the enzyme and nanomaterials. As the distance between the MOFs and the enzyme decreases, the number of H-bonds increases, i.e., the H-bonds number surges with the progress in the simulation. Fig. 6C-ii represents the average H-bonds for each case. A correlation between H-bond number and SASA values seems necessary. With a decrease in the distance between the nanomaterial and the enzyme, the exposed surface area toward water molecules increases. In other words, water molecules between protein and nanomaterial are squeezed out; hence, protein and nanomaterial absorb each other.
Consequently, increases in the H-bonds are observed as a result. Comparing Fig. 6A and C-ii depicts the hypothesized correlation that increased H-bond number results in the decreased available surface area toward solvent molecules. The H-bond number increases with the addition of the hydroxyl functional group that provides more positions for H-bonding with water molecules. By functionalizing the HKUST MOFs with hydroxyl groups, the average number of hydrogen bonds increased from 20 to about 34. Hydrogens attached to the electronegative oxygen atoms provide the condition for forming H-bonds between hydroxy HKUST and Mpro. The formation of these bands makes this interference stronger.
Low values of RMSF and RMSD indicate more stability and balance in the simulation system. The addition of the hydroxyl group has stabilized HKUST MOFs. In addition to having the minimum amount of RMSD, it also has the minimum amount of RMSF. Table 1 shows the average of RMSD and RMSF during the time. As shown in Fig. 4, the best interaction between the enzyme and the pristine MOFs is formed by HKUST. The hydroxyl groups promote the MOFs’ interaction with the enzyme.
Table 1
The average amount of RMSD and RMSF for the control group and MOFs
Structure
|
Average of RMSD (nm)
|
Average of RMSF (nm)
|
Control group
|
5.16
|
8.34
|
ZIF
|
4.85
|
6.17
|
IRMOF
|
4.31
|
3.81
|
HKUST
|
4.11
|
3.74
|
HKUST-OH
|
3.35
|
2.91
|