Apigenin (APG) is a flavonoid found in high concentrations in many fruits, vegetables, and Chinese medicinal herbs. It has a wide range of physiological effects, including anti-inflammatory, antioxidant, antibacterial, anti-viral, and blood pressure-lowering properties (Yan et al., 2017). Apigenin has been shown to have anti-cancer, anti-diabetic, and antioxidant effects. APG inhibits tumor cell growth and survival by triggering apoptosis, cell cycle arrest, and the creation of ROS. It also prevents metastasis and angiogenesis by changing various cellular signaling pathways (Zhou et al., 2022). APG suppresses Enterovirus-71 infection by interfering with RNA-transacting factor interactions (Zhang et al., 2014).

Resveratrol is a natural chemical produced by certain plants (Campagna and Rivas, 2010). Resveratrol has anti-viral activity against human and animal viruses (Abba et al., 2015). Resveratrol has also inhibitory effects on various viruses (Zhao et al., 2017).

Hesperetin is a flavanone, and in its water-soluble glycoside form of hesperidin, it can be found in various citrus fruits. Hesperetin has shown anti-viral properties in vitro against a few RNA viruses (Zandi et al., 2011). Hesperidin also demonstrated anti-viral activity against the growth of the 17D strain of the yellow fever virus (Agrawal et al., 2021). The bioflavonoid Hesperetin inhibits the intracellular reproduction of viruses (Oo et al., 2016).

Fungisterol, a compound found in mushrooms, has anti-viral properties (Zhu et al., 2021). Potential anti-viral effects of Fungisterol have been demonstrated against various pathogens, including the human immunodeficiency virus (HIV), influenza, herpes simplex virus (HSV), hepatitis B and C viruses, and others (Raut and Adhikari, 2021).

Lucidone is a natural compound found in the fruiting body of the edible mushroom Ganoderma lucidum. It belongs to the class of organic compounds known as coumarins, which are fragrant and sweet-smelling organic compounds that are commonly found in plants (Boh et al., 2007). Recent research has been reported that the Lucidone might be composed to have various pharmacological properties, including antioxidant, anti-inflammatory, antitumor, antiviral and neuroprotective activities. It also reported that the Lucidone suppresses dengue viral replication through the induction of heme oxygenase-1 and suppresses Hepatitis C Virus (Chen et al., 2013, 2018).

Ganoderic acid (GA) has been shown to have anti-viral, antihypertensive, anti-cancer, and immune-modulatory effects (Zhu et al., 2021). Anti-viral effects of Ganoderic acid against Enterovirus 71 (EV71) infection are observed without cytotoxicity (Zhang et al., 2014). There is evidence that Ganoderic acid can inhibit viral replication and reduce the extent of liver damage (Li and Wang, 2006).

It has also been demonstrated that Curcumin prevents the development of several viruses and cancerous cells. Curcumin can block the production of HSV-1’s immediate early genes and the action of the HIV-1 integrase, both of which are required for the virus’s replication (Zandi et al., 2010). Through mechanisms present in the infected cells, Curcumin can stop the production of viral genes and, consequently, viral replication (Kutluay et al., 2008).

Previous reports indicate flavanol Kaempferol and its glycosides have strong anti-viral properties (Khazdair et al., 2021). Kaempferol could prevent the spread of viruses in the brain, lungs, kidneys, heart, and spleen. According to research, the hepatitis B virus, the H1N1 and H9N2 influenza viruses, and the Herpes simplex virus are all susceptible to the effects of Kaempferol (Zhang et al., 2012; Rahman et al., 2022).

Since both Cholic acid and metal ions have pharmacological effects, such as antibacterial, anti-viral, antifungal, antimalarial, antitubercular, anti-cancer, spermicidal, and antiallergic, their organometallic complexes were created to have a synergistic effect (Kishu and Siva, 2011). It has been found that several cholic acid derivatives, including taurolithocholic acid, lithocholic acid 3-sulfate, and taurolithocholic acid 3-sulfate, preferentially suppress type 1 human immunodeficiency virus replication (HIV-1) (Baba et al., 1989). Cholic acid can be prevented Coxsackievirus B3 recurrence and produce anti-viral activity (Han et al., 2018).

Chlorogenic acid is a widely distributed natural chemical with numerous significant pharmacological actions, including antioxidant, anti-inflammatory, antibacterial, anti-viral, hypoglycemic, lipid-lowering, anti-cardiovascular, antimutagenic, anti-cancer, and immunomodulatory (Naveed et al., 2018). Chlorogenic acid (CHA) treats viral upper respiratory tract infections brought on by influenza, parainfluenza, and respiratory syncytial viruses (Ding et al., 2017). Besides, another in silico investigation has reported that Chlorogenic acid might be effective against SARS-CoV-2 (Aini et al., 2022).

Pinocembrin demonstrated potent antifungal action in a dose-dependent manner against Penicillium italicum (Peng et al., 2012). Pinocembrin exhibits anti-viral activity against the reproduction of the Zika virus (Ghildiyal et al., 2020; Huang et al., 2022). Pinocembrin inhibits the Zika virus replication cycle’s post-entry process, and also, it also prevents the formation of viral envelope protein and RNA (Guang and Du, 2006).

Sanguinarine is a plant alkaloid with considerable antiangiogenic, anti-cancer, and anti-viral potential but comparatively moderate toxicity (Feldman et al., 2018). Sanguinarine promotes anti-viral activity by upregulating TLR expression and its downstream mediator in MDM (Tilgner and Shi, 2004). It demonstrated selectivity for MKP-1 over MKP-3, inhibiting cellular MKP-1 with an IC₅₀ of 10 m (Guang and Du, 2006).

Below Figure 1 has shown the chemical structures of studied compounds and Table 1 displayed the pharmacological evidence more precisely.

RMSD refers to the occurrence of deviations that were noticed during the evolution of the simulation (Peng et al., 2012; Ghildiyal et al., 2020; Huang et al., 2022). Furthermore, the RMSD defines the stability of the structural as the smaller the RMSD, the greater the stability (Huang et al., 2022). RMSD calculations for protein backbones, ligands, and complexes during a100 ns MD simulation of each protein-ligand complex were used to get an insight into conformational changes during protein-ligand interactions, as shown in Figure 7.

Derived from a 100 ns MD simulation, the RMSD of the ligand, protein backbone, and complex RMSD values for apigenin, curcumin, and Ganoderic acid are displayed. The RMSD of the protein backbone is displayed in red. The complex RMSD (root-mean-square deviation) is depicted in black. The RMSD (root mean square deviation) of the ligand is depicted in green.

Figures 7A,B illustrate that the West Nile virus methyltransferase protein with apigenin ligand has a lower RMSD value compared to the West Nile virus methyltransferase protein with curcumin ligand. In Figures 7A,B, Ligand’s average RMSD is 0.6785 Å and 2.066 Å, respectively. Although the apigenin ligand bound the West Nile virus methyltransferase protein, the low RMSD value showed large fluctuations throughout the simulation. In the case of the ligand curcumin, the fluctuation was just observed during the first 10 ns of trajectory. Furthermore, the ligand RMSD stabilized again after 10 ns and remained consistent till the end of the simulation time of 100 ns. This indicates that the curcumin ligand contacts remained intact during the simulation. According to Figure 7C, the Ganoderic acid ligand-bound West Nile virus envelope glycoprotein showed a stable RMSD value during the simulation time. The trajectories were analyzed with the VMD program, and the results showed that the ligands curcumin and Ganoderic acid did not jump out of the domain of the protein, indicating that they were located within the binding site. This was discovered when we found that neither of the ligands jumped out of the domain. According to the results of the root-mean-square-difference (RMSD) analysis of the ligands, neither the binding orientation of the curcumin nor the Ganoderic acid ligands changed during the simulation. Apigenin and Ganoderic acid have maintained the same conformational orientation throughout their whole structures and have not deviated from them. This suggests that, in the case of the West Nile virus methyltransferase protein-bound Apigenin-ligand, numerous binding orientations were observed and that this ligand shifted positions throughout MD simulations and departed the binding pocket. The docking position was deemed unacceptable for the candidate with a complex structure. During the 100 ns simulation, the RMSD analysis of backbone atoms for the two proteins, West Nile virus methyltransferase and West Nile virus envelope glycoprotein, revealed different conformational states in terms of the protein’s backbone. Indeed, curcumin and Ganoderic acid tend to reach a constant equilibrium, but the RMSD of the apigenin backbone was considerably high. The backbone of Apigenin remained distinct throughout the simulation, resulting in a maximum RMSD of 4.5 Å. This difference in the deviation range explains the change in stability of West Nile virus methyltransferase, which represents the effect of the substituted amino acid on the structure of the protein. The RMSD plots of Apigenin and Curcumin’s backbones have average values of 2.91 and 1.76, respectively. The first 30 ns of the simulation of curcumin showed instability of the protein, but from 30 to 60 ns of simulation, the protein was stable. The first stable conformation of the Ganoderic acid backbone occurs between 20 and 40 nanoseconds, while the second conformation occurs between 60 and 100 nanoseconds. The RMSD remains constant at 1.7 Å, and between 20 ns and 40 ns, RMSD >5 Å is observed to fluctuate significantly. Among the three trajectories, the apigenin backbone protein exhibited the greatest variation and the highest RMSD value. As opposed to Apigenin and Curcumin, the Ganoderic acid protein backbone exhibited less divergent patterns. In addition, according to Figure 7, the average RMSD values for the three complexes were 3.50 Å, 3.97 Å, and 5.50 Å, respectively. As determined by RMSDs, the protein-ligand complex in Figure 7A exhibited a rising fluctuation, whereas in Figure 7B, the complex is equilibrated between 50 and 75 ns. Also, as can be seen in the complicated Figure 7C, the trajectory is initially in equilibrium for 5 ns, then has a big increase to 16 at 20 ns, and finally experiences a substantial reduction to 5.3 at roughly 23 ns. From then on, the complex exhibits a wide range of fluctuations, from nearly stable to extremely unstable, until it reaches 70 ns. In the last 30 ns, the complex has shown persistent, minute fluctuations. As a result, the West Nile virus methyltransferase complex protein-bound apigenin exhibited the lowest RMSD value compared to the curcumin compound and Ganoderic acid compound complexes; however, this did not prove its greater stability and fewer conformational changes compared to other complexes. At the conclusion of the simulation, the ligand appeared highly unstable when complexed with apigenin. The apigenin ligand exhibited substantial variations, indicating the general instability of the complex.

RMSF measures the amount by which atomic locations have deviated from their initial positions and illustrates the dynamic nature of the protein-ligand interaction. In other words, RMSF exemplifies how dynamic the interaction between the protein and the ligand is. As a result, the RMSF data of the protein backbone and complexes were plotted to visualize the average fluctuation of all the amino acid residues, as depicted in Figure 8 for a 100 ns MD trajectory. Furthermore, the RMSF value can be used to assess the significance of individual protein residues in preserving the native shape of a protein-ligand complex. A high RMSF number indicates greater flexibility, while a low RMSF value indicates a more stable zone. Hence, a higher number of RMSF residues or groups suggests a higher degree of flexibility, which in turn suggests a higher probability of interaction with ligand molecules. Moreover, reduced RMSF fluctuations are associated with lesser flexibility, resulting in diminished interaction potential. Figure 8A depicts the RMSF analysis for the ligand apigenin. None of the nine residues with a high RMSF value (RMSF value >3) were outside of protein binding sites. In addition, fluctuations of substantial peaks in the RMSF graph were found at LYS45, GLU46, GLY47, ASN48, VAL49, THR50, GLY51, GLY52, and HSD53, which are shown in blue on the protein. Intriguingly, these residues are not part of the binding pocket, and it is hypothesized that they have no significant effect on the ligand-binding process. The curcumin ligand has four different residues that have high RMSF, but none of them are in the protein binding sites. The Ganoderic acid ligand has not exhibited significant fluctuations with low RMSF values, none of which are in protein-binding sites. As demonstrated in Figure 8A, the RMSF values of the apigenin ligand complex are greater than those of the curcumin structure. The high RMSF values of apigenin indicated a greater degree of flexibility and instability in the protein, whereas the low RMSF values of the curcumin and Ganoderic acid complexes indicated restricted movement of the residues and a rigid structure in the presence of the ligands for West Nile virus methyltransferase. According to the information presented above, the findings of curcumin and Ganoderic acid have demonstrated that, in general, the protein was stable and did not undergo any significant changes. In this investigation, apigenin and protein interactions with RMSF values greater than 2.5 are regarded as having fewer stable bonds. Figure 8, on the other hand, reveals a negligible distinction between the fluctuation scores of the residues in the protein-curcumin combination and the protein-Ganoderic acid complex. In Figures 8B,C, the RMSF plot of ligand-protein for each ligand reveals that the RMSF value at the C-terminal residues (end) is quite high because these residues represent the tails or ends of the protein structure, which are highly reactive and free to move. Each of the two complex system residues, VAL267 (C-terminal), exhibits significantly higher fluctuations than other residues. This is likely because it is placed at the end of the protein chain, allowing it to be more flexible than other residues. In Figure 8C, the C-terminal residue changed significantly, with RMSF values of 60, whereas for curcumin, it reached 30. Figure 8C depicts the creation of a highly mobile free-end loop in Ganoderic acid, which may account for the significant variation observed.

To evaluate structural compression changes, the gyration radius diagram of each structure was recorded over the course of the simulation. To determine the compactness of the system over time, Rg was calculated, with higher Rg values indicating less compactness (more unfolded) with conformational entropy and lower Rg values indicating high compactness with greater structural stability (more folded). Figure 9 demonstrates that the simulation Rg values for the three complexes are 1 Å, 9 Å, 1 Å, 975 Å and 3.5 Å, respectively. From the Apigenin plot, it can be seen that the system exhibited little change until 100 ns, with the exception of a few unfolding events between 40 and 70 ns, after which it stabilized until the end of the simulation. The stability of the protein in the complex was demonstrated by the Rg value, which showed less variation as a result of the fewer changes it experienced. Curcumin, on the other hand, exhibits only very modest fluctuations at the beginning but afterwards demonstrates a significant increase that reaches a Rg value of 2.5 Å between 20 and 40 ns. In the case of Ganoderic acid, the structure showed a sharp increase of Rg up to 3.75 Å at 10 ns, then an immediate decrease of the Rg value towards 3.50 Å, then an increase towards 3.70 Å between 20 ns and 40 ns, then a gradual fluctuation decreases of Rg towards 3.60 Å until the 80 ns. At 80 ns, there is a sharp decrease in the Rg value towards 3.50 Å. After that, an average Rg value of 3.60 A was equilibrated to small fluctuations until the simulation ended. The RG findings demonstrated that the binding of these three molecules induces structural modifications. In general, the patterns of change in RG values across all complexes were distinct. According to the results, the Apigenin complex had the smallest Rg value, which may indicate that it is more compact than the other complexes. A loss of compactness may occur as a consequence of a change in the interaction pattern between the protein and the ligand. The dissolution of hydrogen bonds between molecules might possibly be explained by the curcumin complex’s elevated Rg value. It is also possible that conformational changes in the protein structure were induced by the interaction of curcumin with proteins, which dramatically altered the curcumin complex’s microenvironment. In addition, the Ganoderic acid complex has been reported to have experienced compositional variations, which indicates a less tightly bound structure.

It is worthwhile to investigate hydrogen bonding interactions because they contribute to the binding process and, in particular, persistent hydrogen bonds. Using the VMD hydrogen bond analysis tool, all conceivable hydrogen bonding interactions between the two specified areas, in this case the protein and the ligand, have been explored throughout time. Hydrogen bonds are recognized, and the output contains the overall number of hydrogen bonds as well as their occupancy over time. The ‘Percentage occupancy of the Hbond’ output of the hydrogen bond analysis tool provides access to these values (Figure 10). In docked complexes, significantly more stable hydrogen bonds are reported to be established. During the Molecular Dynamics (MD) simulations, the H-bonds that were present in the docking structures were not only preserved, but additional H-bonds were also found. Apigenin, curcumin, and Ganoderic acid each have their own individual occupancies of identified H-bonds, which are listed in Table 4.

In the case of the apigenin complex, during MD, the protein-apigenin complex stability was maintained by interactions with GLY 81, LYS 105, and ILE 147 residues. It has hydrogen bonds with the highest occupancy of 11.28%, where apigenin forms the hydrogen bond with LYS 105. Then 0.90% occupancy, where apigenin plays a role as an acceptor for GLY 81, and 0.10% with ILE 147. Although the Apigenin complex has 50 hydrogen bonds more than the other complexes, the compound (Apigenin) formed only two hydrogen bonds, which were found in docked simulation. Curcumin has 28 total hydrogen bonds, with the highest occupancy of 80%, where curcumin and VAL 132 form hydrogen bonds. Then 40% occupancy, where curcumin plays a role as an acceptor of ASP 131, and 1.20% with ASP 131. Only three of the 28 hydrogen bonds were formed in the docked simulation. We conclude that, during MD, the curcumin complex has more stability than the apigenin complex. Ganoderic acid has 32 total hydrogen bonds with the highest occupancy of 29%, where Ganoderic acid has the role of a donor and SER 341 has the role of an acceptor of hydrogen bond formation. Then 19% occupancy where Fisetin plays a role as an acceptor to ILE 340 and 19% with PRO 339. Only two of the 32 hydrogen bonds were formed in the docked simulation. In the system, hydrogen bonds were stable, and most of them appeared between 0 and 15 ns and 40 and 50 ns. Hydrogen bond occupancy with a score of more than 100% indicates that more than one atom pair interacts to form hydrogen bonds. Meanwhile, curcumin has a lower total hydrogen bond but a higher hydrogen bond occupancy than Ganoderic acid. This total hydrogen bond and hydrogen bond occupancy determine the stability of each system. The number and occupancy of hydrogen bonds are the keys to the interaction stabilization of the protein-ligand complex.

Principal component analysis was used to identify and understand significant concerted motions in different regions of the protein. PCA extracts the most variable dynamic motions in simulations, which are required for biological function. In addition, it may be used to examine the effect of various parameters on the collective motion and to reduce the motion’s complexity, which is related to the system’s stability and protein functions. Also, it can be utilized to characterize the many conformational variations that are associated with the process of protein folding as well as the open-close mechanism of ion channels. The conformational alterations of West Nile virus methyltransferase and West Nile virus envelope glycoprotein were further analyzed using dynamic cross correlation matrix (DCCM) analysis. Principal component analysis (PCA) and a DCCM were used to identify protein backbone conformation shifts. PCA and DCCM were performed using RStudio and Bio3d; axes, each dot represents a different protein configuration. The distribution of blue and red dots represented the extent of conformational changes in the simulation, where the color ranges from blue to white to red corresponds to simulation duration. The color blue represents the initial timestep, the color white represents the intermediate, and the color red represents the final timestep (illustrated in Figure 11).

The first three PCs predicted the majority of the motion of the protein backbone from the MD trajectories in Figure y. In the apigenin protein, PCA analysis shows that the first three eigenvectors account for 47.49, 47.49, and 7.79% (Figure 12A). In the case of curcumin protein, PCA analysis display that the first three eigenvectors account for 23.79, 23.79, and 6.13% (Figure 12B). For Ganoderic Acid Protein, PCA analysis display that the first three eigenvectors account for 30.1, 30.1, and 11.34% (Figure 12C). The highest PC1 (47.49%) was noticed for the apigenin protein, which indicates that the protein had undergone higher conformational changes. The lowest PC1 (6.13%) was observed for the curcumin protein, indicating that the protein had undergone very few conformational changes compared to apigenin. Comparatively, the Ganoderic acid compound exhibited less conformational change than the apigenin complex but more than the curcumin complex.

Additionally, the resulting dynamical cross-correlation graphs exhibit both positive and negative amino acid correlation effects. Overall correlation was displayed by DCCM, and it was in the range of −1.0 to 1.0. (From dark purple to dark blue). Various colors were used to denote varying degrees of association between residues, with darker colors indicating stronger correlations. Correlations closer to 1 indicated that the residues were moving in the same direction, while correlations closer to −1 indicated that the residues were moving in the opposite direction. In order to see the relationship between the I and J residue indices, we constructed pairwise correlated graphs. Colors such as dark cyan, white, and pink were used to analyze the predicted map results. Fully correlated pairs are denoted by the color cyan, while anti-correlated pairs are denoted by the color pink. Comparative results reveal that the atomic motions in the Apigenin complex resemble more closely those of the Curcumin structure. But, Comparing the DCCM diagrams of the three systems, it could be found that the correlated motions of the apigenin and curcumin systems were noticeably distinct from the Ganoderic complex. We postulate that the simultaneous appearance of positively and negatively correlated movements destabilizes the domain. For this reason, we observed high anti-correlation in the apigenin complex compared to the curcumin complex, indicating a more compact structure of the apigenin. For the Ganoderic acid complex, which showed less anticorrelated and correlated motions and more noncorrelated motions than the Apigenin and Curcumin complexes in all regions of the protein (illustrated in Figure 12).

To analyze the molecular binding interaction of protein-ligand complexes, the binding free energy (G) was calculated using MM-PBSA, which considers both bonded and non-bonded (van der Waals and electrostatic) interactions. Using MMPBSA and the final 20 ns of the trajectory, the binding free energies of the protein-ligand complexes were calculated. Using the MMGBSA approach, the binding free energy (∆ Gbind) of 12 natural compounds, of which top nine were chosen based on binding affinity score, was determined. The greater the negative values, the more favorable the binding free energy between proteins and ligands. As shown in Table 4 and Figure 13, the free binding energies of the natural compounds agree with the molecular docking results. The Hesperetin, Lucidone, and Cholic acid demonstrated the greatest amount of binding energy (−16.44 kcal/mol, −19.78 kcal/mol and − 29.00) in comparison to the other natural chemicals. This energy is mostly different from the curcumin compound, and it means that the interaction in the compound Hesperetin, Lucidone, and Cholic acid complex is the most stable interaction during simulation, followed by the protein-compound complex. On the other hand, the compound Ganoderic acid complex has the most positive binding energy, indicating that the interaction between compound Ganoderic acid and the target protein active site is the weakest.

Table 5 and Figure 13 highlights the primary components of binding free energy for protein-ligand complexes, including van der Waals (ΔEVDW), electrostatics (ΔEEEL), a polar portion of solvation (ΔGPB), non-polar part of solvation (ΔGNP), dispersion (ΔGDISP), and binding energy (ΔG).