Malacitanolide, reissantin E and paclitaxel compounds as inhibitors of envelope, NS5 and NS2B/NS3 target proteins of dengue virus: Computational docking and molecular dynamics simulations studies

1 Introduction

Dengue is one of the important mosquito-borne diseases which is caused by four serotypes of dengue virus (DENV 1–4) (Bhatt et al., 2013). Aedes aegypti and Aedes albopictus are the major insect vectors that transmit dengue virus to human beings. Dengue haemorrhagic fever (DHF) is an important clinical symptom of dengue viral disease (Wang et al., 2020). World Health Organization estimated that around 390 million dengue cases occur annually and around 96 million cases have disease severity. The genome of dengue virus contains a positive sense long Open Reading Frame (ORF) encoding a polyprotein that contains structural proteins [C (capsid), prM (membrane), and E (envelope)] and non-structural proteins [NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5] (Nowak et al., 1989).

The E protein is involved in the attachment of virus to the host cell and fusion of virus to host cell membrane (Perera and Kuhn, 2008). NS2B/NS3 proteins process the polyprotein precursor prior to the assembly of viral replicase complex (Wahaab et al., 2021). The NS5 has RNA methyltransferase (MTase) and RNA-dependent RNA-polymerase (RdRp) in the N-terminal and C-terminal regions, respectively (Rodenhuis-Zybert et al., 2010). Based on the above vital functions, E, NS5 and NS2B/NS3 proteins are considered as potential targets for finding out antiviral therapeutics (Wahaab et al., 2021).

Target-based drug identification is mainly used for drug discovery (Lim et al., 2013). Phytocompounds are mainly targeted for drug development against many viral diseases. The important phytocompound groups are flavonoids, terpenoids, alkaloids, and steroids (Othman et al., 2019). Many studies reported that flavonoid derivatives are potential inhibitors against the dengue virus by inhibiting its replication (Allard et al., 2011; Frabasile et al., 2017). Phytocompounds belonging to the terpenoid group are studied extensively against the dengue virus. Therefore, in the present study, terpenoid compounds were studied for their ability as potential inhibitors against target proteins of dengue virus using in silico methods.

2 Materials and methods

3 Results

In the present study, a total of 513 terpenoid compounds were screened against three target proteins using iGEMDOCK. Top 50 compounds showing higher binding energy values were analysed for accurate docking and drug screening. Then, best ten terpenoid compounds such as malacitanolide, reissantin E, paclitaxel, gamma-tocopherol, saikosaponin B2, brunneogaleatoside, taiwaniaquinol A, englerin A, taxumairol, and saikosaponin D were docked with target proteins by Molecular Docking Server. It was observed that the phytocompounds paclitaxel, reissantin E, and malacitanolide exhibited docking scores of −6.43, −6.34 and −6.29 kcal/mol, respectively against 1OKE target (Fig. 1). The reference ligand BOG exhibited the interaction energy value of −5.70 kcal/mol against 1OKE. Likewise, the interaction energy values of −9.39, −7.56, −6.12 and −5.14 kcal/mol were obtained for paclitaxel, malacitanolide, reissantin E and SAH respectively against 1R6A-SAH, values of −7.37, −6.52, −6.19 and −5.90 kcal/mol for paclitaxel, malacitanolide, reissantin E and RVP respectively against 1R6A-RVP and values of −6.49, −6.43, −5.83 kcal/mol for paclitaxel, reissantin E and malacitanolide respectively against 4M9K. An inbuilt reference ligand was not reported for the 4M9K hence binding site region suggested by the MDS was adopted for docking analysis (Table 1).

Close

Fig. 1

Chemical structures of reference ligands and terpenoid compounds.

Export to PPT

3.1

3.1 Induced Fit docking Maestro 10.7

The screened compounds were further subjected to Maestro 10.7. Against 1OKE, malacitanolide showed Glide GScore of −7.02 kcal/mol and XP HBond of −3.056 kcal/mol and formed HB interactions with Thr280, Gln200, Gln271, Gln49 and Ala50 (Figs. 1 & 2), while paclitaxel showed Glide GScore of −6.895 kcal/mol and XP HBond of −2.125 kcal/mol and formed HB interactions with Glu49, Lys128, Asp203, Thr48 and Ala50 and reissantin E showed Glide GScore of −4.426 kcal/mol and XP HBond of −1.220 kcal/mol and formed HB interactions with Ser274 and Gln200. Against 1R6A-SAH, paclitaxel showed Glide GScore of −9.334 kcal/mol and XP HBond of −3.367 kcal/mol and HB interactions formed with Lys61, Asp146, Trp87, Gly148, and Arg84, while Malacitanolide showed Glide GScore of −6.130 kcal/mol and XP HBond of −2.055 kcal/mol and formed HB interactions with Lys105, Gly106, Glu149, and reissantin E showed Glide GScore of −5.840 kcal/mol and XP HBond of −2.105 kcal/mol and formed HB interactions with Ser150 and Lys22 (Figs. 2–4).

Close

Fig. 2

The 2D interaction plots of compound Malacitanolide with the BOG site of envelop protein of dengue virus.

Export to PPT

Close

Fig. 3

The 2D interaction plots of compound Malacitanolide with the 1R6A-RVP site of dengue virus.

Export to PPT

Close

Fig. 4

Docking pose of compounds (a) Paclitaxel and (b) Reissantin E with the 1OKE target protein of dengue virus.

Export to PPT

While malacitanolide showed Glide GScore of −5.295 kcal/mol and XP HBond of −1.625 kcal/mol and formed HB interactions with Ser150, Lys14 and Lys22, Paclitaxel showed Glide GScore of −4.561 kcal/mol and XP HBond of −1.862 kcal/mol and formed HB interactions with Lys22, Lys29 and Ser214 and reissantin E showed Glide GScore of −3.363 kcal/mol and XP HBond of −1.438 kcal/mol and formed HB interactions with Lys105, Gly148 and Glu149 against 1R6A-RVP. Against 4M9K, the malacitanolide showed Glide GScore of −4.030 kcal/mol and XP HBond of −1.145 kcal/mol and formed HB interactions with Glu1169 and Asn1119, whereas reissantin E showed Glide GScore of −3.943 kcal/mol and XP HBond of −1.460 kcal/mol and formed HB interaction with Asn 1119 and paclitaxel showed Glide GScore of −2.034 kcal/mol and XP HBond of −1.008 kcal/mol and formed HB interaction with Arg1157 (Figs. 2 & 3). The Glide Energy and the interactions of other non-bonded amino acids of the above three compounds with protein targets using Maestro 10.7 (Schrödinger) are given in Table 1.

3.2

3.2 Molecular dynamics (MD) simulation

Structural dynamics of malacitanolide bound to 1OKE were analysed by performing molecular dynamics simulations for 5 ns. The top-ranked compound with IOKE of dengue virus was chosen for MD simulations. All five properties, like Energy fluctuations, RMSD, Root Mean Square Fluctuation (RMSF), ROG (radius of gyration), and H-bonds were analysed for assessing the protein–ligand complex stability. RMSD of protein structure backbone was simulated by plotting a graph with nanosecond verse angstrom units. The protein fluctuation was greatly observed until 1.5 Å at the initial stage, where slight conformational changes occurred after which the system attained stability. The Fig. 5 explains the stability of the complex and the degree of fluctuation. The RMSF of carbon α residues of the protein chain fluctuated at specific range of amino acids viz. 53–90, 101–110, and 325–329. These fluctuating sites mainly formed turns and coils, not the alpha helix or beta sheets. The protein residues involved in ligand contacts were highlighted in green colour. The ligand had maintained a stable confirmation at 1.5 to 2.0 Å. Initially, the fluctuation was seen in ROG for the ligand until 1 ns, then acquired and showed a regular conformation throughout the simulations. The radius of gyration ranged from 4.0 Å to 4.16 Å throughout the 5 ns simulations, and while looking into the interaction profile of ligand during simulation we observed that Thy280 had formed two hydrogen bonds with malacitanolide with 95% and 67% stability and Ala50 formed one hydrogen bond with 81% stability. A few other amino acids such as Glu49, Lys128, Thr48 and Lys47 also contributed to interactions supported by water bridges and this finding supports that water molecules maintaining the receptor–ligand complex (Fig. 4).

Close

Fig. 5

Molecular dynamic simulation of Dengue virus 1OKE complexed with malacitanolide.

Export to PPT

4 Discussion

The unavailability of effective antiviral drug in treating dengue infections complicates the clinical management of dengue patients. This study analyses the terpenoid compounds against dengue viral target proteins such as structural E protein and nonstructural proteins like NS5 and NS2B-NS3 using computational approaches. Many studies have suggested that the dengue E, NS2B/NS3 and NS5 proteins are potential targets to identify antiviral molecules (Modis et al., 2003; Hrobowski et al., 2005; Tomlinson et al., 2009; Murugesan and Manoharan, 2020). Non-structural proteins were also the most important targets due to their involvement in the dengue virus replication. They may have two domains of conserved nature, namely, methytransferase and RdRp (Rawlinson et al., 2006). It was found that among the terpenoid compounds screened in this study, three compounds contributed to the effective interactions with the structural and non-structural protein targets of the dengue virus. A study reported that the compounds doxycycline and rolitetracycline have shown an effective interaction with the envelope protein and exhibited hydrogen bonds with amino acids such as Thr48, Gln49, Ala50, Lys51, Glu52, Gln271 and Gln200 in the BOG site and also showed antiviral activity at 67.1 µM and 55.6 µM, respectively (Yang et al., 2007). In this study, malacitanolide formed the hydrogen bond with amino acids such as Thr280, Gln200, Gln271, Gln49 and Ala50 in the BOG site and it was hypothesised that malacitanolide compound would have significant antiviral properties against dengue virus owing to similar interaction profile with that of doxycycline and rolitetracycline.

It was noted that malacitanolide formed five HB interactions with both hydroxyl and ketone moieties. The amino acids Thr280, Glu49, Aln50 and Gln200 formed the HB interactions. Likewise, another lead compounds, namely, paclitaxel formed 5 HB interactions and reissantin E formed 2 HB interactions with E protein target. HB interactions of the screened lead compounds against E protein observed in this study were higher than the flavonoid compounds reported by Mir et al. who found that baicalein exhibited HB interaction with Thr48, fisetin with Thr48, Ile270 and Thr280, quercetin with Thr280 and naringenin did not form any HB interaction (Mir et al., 2016).

The docking results revealed that malacitanolide exhibited high interaction energy (-5.295 kcal/mol) with NS5 protein and formed HB interactions with hydroxyl and ketone moieties. Lim et al. (2013) found that the compound SPH1-103–799 formed HB interactions with Lys22, Leu20 and Asn18 of NS5 and the compound SPH1-101–102 formed HB interactions with Lys14, Ser150 and Lys22 of NS5. In this study, the amino acids such as Lys-14, Leu-17, Ser-150, Ser-151, Asn-18, Leu-20, Phe-25, Lys-29, Pro-152, Glu-408, and Gly-409 were significantly involved in the ligand binding with NS5. All the above mentioned amino acid interactions were agreed with the findings of Benarroch et al. (2004) who reported that the same amino acid interactions in RNA cap site except Glu-408 and Gly-409.

Another study identified few significant interactions of SAH with the SAM site of NS5 protein. The amino acids such as Lys105, His110, Asp131 and Ile147 were involved in the interactions and methylation process (Dong et al., 2010). We found that malacitanolide developed hydrogen bond interactions with Lys105, Gly106, and Glu149 and other interactions with Ile147, Asp131, and His110 of NS5 protein. The paclitaxel formed a hydrogen bond with four hydroxyl moieties in NS5 (1R6A-SAH) and the two ketone moieties contributed two hydrogen bonds. Paclitaxel also formed a pi-pi interaction with Arg57 and Arg84 and a hydrogen displacement site was seen at Lys105.

Lim et al. (2013) also reported in another study that the compounds SH1-007–088 and SPH1-111–460 formed hydrogen bond interactions with Ser56, Gly86, Asp146 and Lys105. Another study by Singh et al. (2016) reported HB interactions of analog 1 with Lys181 and Arg57 analog 15 with Ser150 and analog 17 with Lys181 and Arg84.

Against NS2B/NS3 (M49K), the malacitanolide showed docking score of −4.327 kcal/mol. The amino acids such as Arg157, Lys117, Val154, Thr156, Gly153, Val155, Lys73, Ile123, Thr118, Thr120, Ser171 and Glu169 were involved in the above interactions. The other two molecules reissantin E and paclitaxel failed to form sufficient binding energy. It was noted that all three compounds interacted with the different active sites. However, the hydroxyl and ketone groups of the terpenoid compounds contributed to the hydrogen bond interaction. Existence of oxygen functional group increases the biological activities of terpenoid compounds (Wu et al., 2015). Similarly, the hit compound in this study also possesses the hydroxyl and ketone moieties at different positions.

Molecular dynamics simulations have played a significant role in developing antiviral drugs by showing the importance of residual motion in the binding of ligand molecules. Molecular dynamics simulations could provide the function of the time of motion of an individual particle (Karplus and McCammon, 2002). In the MD simulations studies, the fluctuations are required for protein–ligand interaction and conformational changes in the protein. MD simulation was reported for less than 10 ns in earlier studies and now which has increased to 100 ns due to the availability of advanced computers. The temperature of the protein and the solvent used as either 180 K or 300 K. In a study, various bioflavonoid compounds were studied for interactions with dengue-RdRp target. Protein-ligand interaction of hesperidin was confirmed in simulation analysis for 5 ns and identified as potent RdRp inhibitor (Fatriansyah et al., 2022). In another study, mulberroside A chebulic acid, mulberroside C and curcumin were identified as potent anti-dengue virus phytocompounds by MD simulations studies (Vora et al., 2019). Similarly, in our study, we performed the MD simulations at temperature of 300 K and at time of 5 ns. MD simulations proved effective in visualising the migration of water molecules and stabilising the protein–ligand interaction. Finally, while looking into the quantum chemistry aspect of the studied compounds, all three compounds had very low band gap energy (ΔE) between both HOMO and LUMO, favouring the transfer of energy and exchange of electrons and would relatively result in increased activity of the compounds.

5 Conclusions

The screened terpenoid compounds, viz. malacitanolide, reissantin E, and paclitaxel were identified as potential inhibitors against envelope, NS5 and NS2B/NS3 target proteins of dengue virus. Moreover, these compounds possess good pharmacokinetic properties that would support their usage as therapeutics for dengue. Both in-vitro and in-vivo studies have to be conducted to further confirm the anti-dengue virus activity of the above three screened terpenoid compounds.