3.6.  Molecular dynamics (MD) simulation study
The MD simulation is one of the proven in-silico methods for the determination of protein–ligand dynamics concerning a high temporal resolution of nanosecond or picosecond order. Here the docked poses of RBD Spro and Mpro with Piperine were used for a 100 ns MD simulation to analyse the stability of these docked compounds.

3.6.1.  Root mean square deviation (RMSD)
The RMSD values from MD simulation provide information about structural and conformational stability. Figure 5 represents the backbone RMSD data of viral proteins and their complex with Piperine. From the plot, it is observed that both the simulations have less fluctuation throughout the simulation time. The average RMSD values of RBD Spro, RBD Spro-Piperine, Mpro and Mpro-Piperine are calculated as 0.143 ± 0.025 nm, 0.130 ± 0.018 nm, 0.212 ± 0.041 nm and 0.203 ± 0.028 nm, respectively. The average RMSD values of the Piperine bound proteins as compared to only proteins are found to be less representing to their conformational stability. Both the simulations are attained equilibrium within 0.3 nm, which is also a measure of the systems’ stability during the simulation (Al-Shabib et al., 2018, 2020; Millan et al., 2018).
Figure 5.  Root mean square deviation plots of (a) RBD Spro (black) and RBD Spro-Piperine (red); (b) Mpro (black) and Mpro-Piperine (blue).

3.6.2.  Root mean square fluctuation (RMSF)
The conformational fluctuations of the proteins RBD Spro and Mpro were analysed by observing the residual changes that resulted due to the interaction of Piperine with the proteins. The RMSF plots of the Cα atoms of the viral proteins and their complex with Piperine are shown in Figure 6. From the analysis, it is found that RBD Spro, RBD Spro-Piperine, Mpro and Mpro-Piperine have the average RMSF values 0.099 ± 0.060 nm, 0.097 ± 0.051 nm, 0.119 ± 0.077 nm and 0.120 ± 0.077 nm, respectively. It is observed that RBD Spro-Piperine (Figure 6(a)) and Mpro-Piperine (Figure 6(b)) show similar fluctuations as compared to only RBD Spro and Mpro, which implies to the stability of these compounds. In addition to that, a majority of the protein residues are found to be stabilized within RMSF 0.3 nm. The decrease in fluctuations of Piperine bound to RBD Spro also suggests for the active binding of Piperine (A. Kumar, Choudhir, et al., 2020).
Figure 6.  Root mean square fluctuation plots of Cα-atoms of (a) RBD Spro (black) and RBD Spro-Piperine (red); (b) Mpro (black) and Mpro-Piperine (blue).

3.6.3.  Radius of gyration (Rg)
The root mean square distance between an object and the centre of gravity is defined as the radius of gyration (Rg). The radius of gyration is a measure of the compactness of the protein structure, where higher Rg value is referred to as a less compact structure, and low Rg value is inferred as high compactness that implies more stability. The measured average Rg values of RBD Spro, RBD Spro-Piperine, Mpro and Mpro-Piperine are 1.829 ± 0.008 nm, 1.833 ± 0.010 nm, 2.233 ± 0.012 nm and 2.238 ± 0.013 nm, respectively. From Figure 7(a,b), it is observed that there is a little enhancement in the Rg values of RBD Spro-Piperin, and Mpro-Piperine as compared to RBD Spro and Mpro, which implies to the gain in compactness of the protein structures upon binding to Piperine.
Figure 7.  Radius of gyration plots of (a) RBD Spro (black) and RBD Spro-Piperine (red); (b) Mpro (black) and Mpro-Piperine (blue); (c) Intermolecular hydrogen bonds formed between RBD Spro-Piperine and Mpro-Piperine during 100 ns MD simulation.

3.6.4.  Number of hydrogen bonds
The number of hydrogen bonds formed between the protein–ligand complex is the measure of the binding strength of the ligand to the protein. The RBD Spro (red) and Mpro (blue) bound to Piperine have a constant number of 1–2 hydrogen bonds throughout the simulation time (Figure 7(c)). There is a maximum number of 3 and 4 hydrogen bonds observed in the case of RBD Spro-Piperine and Mpro-Piperine, respectively. The number of hydrogen bonds fluctuates throughout the simulation time for both RBD Spro-Piperine and Mpro-Piperine, which suggests for conformational changes in the binding site of the ligand during the simulation. The observation from hydrogen bond analysis indicates that the complexes are stable for the performed simulation time.

3.6.5.  Interaction energy
The interaction energy is the measure of the interaction strength of the protein–ligand complex. In order to validate the results of molecular docking studies, the analysis of the interaction free energies from MD simulation was performed. The average interaction energy takes the contribution from the average short-range Lennard-Jones and van der Waals energy. The average interaction energies of RBD Spro-Piperine and Mpro-Piperine are found to be − 41.401 ± 17.843 kJ/mol and −143.162 ± 23.043 kJ/mol, respectively. These interaction energy values suggest that Piperine binds to the RBD Spro and Mpro with good affinity and hence supports the docking results, which in turn helps for the favourable use of Piperine as a drug candidate for SARS-CoV-2.

3.6.6.  Solvent accessible surface area (SASA)
SASA is a measure of the receptor exposure to the solvent environment during the simulation. The hydrophobic residues that got exposed to the solvent environment upon binding with the ligand molecules contribute to the SASA values. The plot of the SASA for the proteins and their ligand-bound form is presented in Figure 8. The analysed average SASA values for the RBD Spro, RBD Spro-Piperine, Mpro and Mpro-Piperine are 106.976 ± 1.602 nm2, 107.235 ± 1.667 nm2, 150.698 ± 2.565 nm2 and 151.022 ± 2.207 nm2, respectively. There is no significant change observed for the averaged SASA values of the complex as compared to only protein suggesting their stability after binding to the drug molecule.
Figure 8.  Solvent accessible surface area (SASA) plots of (a) RBD Spro (black) and RBD Spro-Piperine (red); (b) Mpro (black) and Mpro-Piperine (blue).

3.6.7.  MMPBSA binding free energy analysis
MD simulation can also be used to calculate the binding free energy of the protein–ligand complex. The binding free energy is the measure of the stability of the system in turns of consistency of nonbonded interactions throughout the simulation. The binding free energy was calculated by using MMPBSA method by taking 2000 snapshots from the trajectory. The computed value of binding free energy for RBD Spro-Piperine is found to be −5.533 ± 0.839 kJ/mol, and for Mpro-Piperine is −37.971 ± 0.271 kJ/mol. It is observed that for both RBD Spro and Mpro, van der Waals energy plays a crucial role in the interaction process. The van der Waals energy, electrostatic energy and non-polar energy are contributed actively to the total interaction energy. In contrast, polar energy has a positive contribution to the whole interaction process. The observed data indicate that the van der Waals, electrostatic and non-polar interactions combinedly contribute to the stability of both the compounds. The contribution from different interactions to the binding free energy for RBD Spro-Piperine and Mpro-Piperine is provided in Supplementary Table S2.
Table 2.  Lowest energy binding affinity of Piperine and few of the currently used drugs for SARS-CoV-2 as obtained from molecular docking study.
Molecule  Binding affinity (kcal/mol)
RBD Spro  Mpro
Piperine  –6.4  –7.3
Chloroquine  –5.0  –4.9
Favipiravir  –5.3  –5.6
Hydroxychloroquine  –4.8  –6.0
Oseltamivir  –5.1  –5.5
Remdesivir  –6.1  –7.2
Ribavirin  –5.6  –6.1

3.6.8.  Principal component analysis (PCA)
The PCA is an essential technique to monitor the conformational dynamics of biomolecules. It is useful in determining the concerted motion of protein as well as protein–ligand complex from the MD trajectories. The diagonalization of the covariance matrix of backbone atoms of the proteins and ligand-bound form were considered for the principal components PC1 and PC2 (Figure 9). From Figure 9(a,b), it is observed that both the Spro-Piperine and Mpro-Piperine are less flexible as compared to unbound proteins since they covered less conformational space. It concludes that the ligand-bound forms are more stable as compared to the unbound proteins.
Figure 9.  Principal component analysis of (a) RBD Spro (black), RBD Spro-Piperine (red) and (b) Mpro (black) and Mpro-Piperine (blue). Free energy landscape plot of (c) RBD Spro, (d) RBD Spro-Piperine and (e) Mpro and (f) Mpro-Piperine.
The principal components obtained were used as the reaction coordinates to find the Gibbs free energy landscape (Figure 9) to visualize the energy minima of the unbound protein as well as the protein–ligand complex. From Figure 9(c–f), it is observed that both the ligand-bound proteins have less Gibbs-free energy values than the unbound proteins indicating their stability and energetically favourable conformational transitions. The shape and size of the minimum energy area (blue colour) in case of RBD Spro-Piperine and Mpro-Piperine are more as compared to the unbound proteins RBD Spro and Mpro, which suggests the ligand-bound forms are thermodynamically more favourable.
The comprehensive study reveals that Piperine forms a stable complex with RBD Spro and Mpro and can be considered as an active inhibitor against SARS-CoV-2. From the docking results, it is observed that the Piperine molecule is the best candidate for the inhibition of the RBD Spro and the Mpro of SARS-CoV-2 among the selected 30 molecules. To observe the effectiveness of Piperine over currently used drugs, we carried out the docking study of a few drug molecules such as chloroquine, favipiravir, hydroxychloroquine, oseltamivir, remdesivir and ribavirin using the same docking protocol as followed for the 30 spice molecules. From the docking score, it is found that Piperine performed better as compared to the currently used drugs stated above. The lowest energy pose of a few presently used drugs with their 2D interaction diagram is provided in Supplementary Figures S5 and S6 corresponding to SARS-CoV-2 Mpro and RBD Spro, respectively. A comparison of the lowest energy dock scores of these drug molecules along with Piperine is also provided in Table 2. The MD simulation results reveal that Piperine actively inhibits both the RBD Spro and Mpro by binding to their active sites. Piperine binds on the active site of the RBD Spro with those residues by which it interacts with ACE2. So, the binding of Piperine on that site may potentially cease the interaction tendency of RBD Spro with ACE2. Similarly, the interaction of Piperine on the active site of the Mpro may inhibit its viral replication. From the docking and MD results, we conclude that Piperine forms a very stable complex with RBD Spro and Mpro and shows better affinity as compared to the currently used drugs that are mentioned above against SARS-CoV-2.