In Silico Analysis of Active Compounds from Allium tuberosum as Drug Candidate for Inhibitor DENV-3 Envelope Protein

Dengue fever has become a global health issue, the development of dengue vaccine has not yet been established. Medicinal plants are an ideal alternative for DENV infection drugs. The purpose of this study was to determine in silico the potential of active compounds from Allium tuberosum as envelope protein inhibitors of DENV-3. The method of this research is to do docking analysis of compounds with DENV-3 envelope protein and analysis of amino acid residues using MVD, pharmacokinetic analysis using SwissADME, toxicity analysis using ProTox-II. The best docking value for the potential activity to inhibit the receptor DENV-3 is the thymidine compound (RS: - 81.1245 kcal/mol). The highest activity of thymidine is the most promising as a drug candidate, as evidenced by the toxicity analysis which is predicted to have non-carcinogenic, non-mutagenic, inactive properties against hepatotoxicity, cytotoxicity and immunotoxicity parameters, as well as pharmacokinetic analysis that fulfills 6 parameters of lipopolicity, molecular weight, polarity. , insolubility, insaturation, and flexibility which indicate the drug candidate of thymidine is safe for its bioavailability. The conclusion from the results of this study is that one compound has the ability as an antiviral, binding score with DENV-3 is good, and is safe in terms of pharmacokinetics and toxicity, namely thymidine compound.

The dengue disease is caused by the family Flaviviridae which has 4 different serotypes.Dengue virus serotypes are from DENV 1, DENV 2, DENV 3, DENV 4 (Costa et al., 2012).According to Samsono (Sasmono et al., 2019) The incidence of dengue virus infection in Indonesia is generally caused by dengue virus serotype 3 (DENV3).The DENV genome is 10.7 Kb in size with single-stranded positive-sense RNA, encoding three structural proteins namely Capsid (C), Membrane (M), Envelope (E), and seven non-structural proteins (NS1 -NS7) (Haryanto et al., 2016).Both structural proteins and non-structural proteins have an important role in the process of infection into human cells, so they are generally studied in the search for DENV vaccine candidates.
However, the development of the DENV vaccine has not yet been established, due to the variety of DENV serotypes (Messina et al., 2014).In addition to vaccine development research, currently research on anti-DENV drug discovery is also being developed.The development of the search for drug candidates was carried out by preliminary testing in the form of an in-silico test of active compounds with viral ligands by bioinformatics.
Allium tuberosum contains various bioactive compounds, including adenosine, thymidine and tuberosine (Jannat et al., 2019).Allium tuberosum has been identified as having antiinflammatory, antiviral and immunomodulatory activities, its bioactive compounds have shown promising properties in various studies, further research which are promising for development as a DENV drug candidate (Nhut et al., 2020).So far, adenosine, thymidine and tuberosine have not been widely studied regarding their activity as candidate virus inhibitors, especially DENV.This study aims to analyze tuberoside as a DENV drug candidate in silico.
In this in silico study, the viral ligand that will be used is the DENV E glycoprotein complex.Glycoprotein E plays a role when the virus will enter the host cell by binding to the host receptor (Routhu et al., 2019).Furthermore, there will be a conformational change in the host membrane due to a decrease in the pH of the endosome, which will lead to fusion of the virus with the host cell membrane.The selection of the DENV E protein was based on the important fact that it is a key protein before the virus fuses.The hope is that with the (Nuryady, et al) interaction of the active compound tuberoside with the glycoprotein E DENV complex, it will prevent viral infection to host cells .

Ligand structure preparation
Compound IUPAC name information on Allium tuberosum obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov).Next, prepare the ligand by drawing a 2D structure with ChemBioDraw Ultra 16 software.Each of the above structures is then converted into 3D using the ChemBio3D Ultra 16 program.Then energy is minimized using the M erck M olecular Force Field 94 (M M FF94) method on the ChemDraw application.

Preparation of protein structure E DENV-3
Figure 1.Crystal structure of the dengue virus serotype 3 envelope protein domain III The structural protein structure of DENV-3 in .pdbdocument format was downloaded via the RCSB macromolecular biology structure information portal (https://www.rcsb.org).The downloaded protein is the DENV-3 envelope protein (PDB ID: 3UZE) (Hadi & Hamzah, 2013).

Molecular docking
The docking process begins by entering the compound preparation results and the DENV-3 E protein preparation results into the M olegro Virtual Docker (M VD) software (CLCbio, 2013).At this stage, the removal of water molecules (H2O) around the protein structure is also carried out, it aims so that the water molecules do not interfere with the docking process, so that it can be ascertained that only ligands interact with proteins, then cavity detection, validation of the docking method and then docking of test compounds with receptors are carried out target (Zaidan et al., 2019).

Prediction of compound toxicity
Toxicity test was carried out using the ProTox-II website (https://toxnew.charite.de/-protox_II/),using canonical SM ILES the compounds would be processed for their toxicity properties.

Pharmacokinethic analysis
The pharmacokinetic analysis was carried out through the SwissADM E website which can be accessed via (http://www.swissadme.ch/).The first step was to copy the canonical SM ILES of the three active ingredients, and paste them in the SwissADM E instruction column.The pharmacokinetic/bioavailability analysis is processed when clicking the run button

RESULTS AND DISCUSSION
Cavity Determination and Receptor Validation Cavity detection aims to predict areas that have the potential as active sites for proteins to bind to ligands (Pratoko, 2012).The active site protein is green, and the Cavity used is Cavity 5 Vol 33.28 Surface = 102.4(Figure 2).The validation of the docking method was carried out by re-docking the native ligand with the envelope protein DENV-3 PDB ID 3UZE (Table 1).The parameter used to provide an assessment of the validity is by looking at the RMSD (Root Mean Square Deviation) value.RMSD is a parameter that describes how big the change in the interaction of the ligand surface on the crystal structure of the protein before and after docking (Nursanti et al., 2022).The docking method is said to be valid if the RMSD value is < 2 so that it can be used for docking the test compound.The smaller the resulting RMSD value indicates that the predicted ligand position is getting better (Puspaningtyas, 2013).other tests, because it has the lowest energy value so that the protein-ligand bonding ability is more stable (Wei & Cai, 2008).The results of the rerank score (RS) describe the binding of the ligand to the receptor, the rerank score can be used to evaluate the quality of docking, predict and measure the value of binding affinity, and can find the right ligand conformation by looking at the lowest value.Based on the RS, it is possible to predict the activity of the test compound when it has bound to a protein, the lower the RS the more stable the bond between the protein-ligand so that the activity of the compound can be said to be higher and vice versa.(Zaidan et al., 2019).3D visualization of ligand-receptor interactions can be seen in (Figure 3).

Figure 3. 3D visualization of ligand-receptor (3UZE) interactions
The results of molecular docking between the ligand and the receptor can also be seen in the interaction of the resulting amino acid residues so that we can find out more about the analysis of molecular docking and can determine whether the compound has the same inhibitory activity as the native ligand (Schaeffer, 2008).The thymidine test compound produced the same amino acid residues as the native ligand, namely Tyr 81, Arg 20, Thr 72, Thr 72 (Table 2).According to (Syahputra et al., 2014) The same amino acid residue as the native ligand indicates that the ligand is able to inhibit the activity of the target protein and has the potential to have the same function as the native ligand.Thymidine compounds also have more hydrogen bonds than the original protein ligands, this illustrates that the more hydrogen bonds between molecules, the more stable the bonds (Syahputra et al., 2014).2D visualization of ligand-receptor interactions can be seen in (Figure 4).

Prediction of compound toxicity
Toxicity prediction aims to determine the nature of a candidate drug compound along with the possible risks that arise from these compounds and can have an impact on humans (Sqra, 2022).The results of the toxicology prediction (Table 3) with the parameters of hepatotoxicity, carcinogenicity, immunotoxicity, mutagenicity, and cytotoxicity showed that the compounds studied were thought to be non-carcinogenic, non-mutagenic, meaning that the compound was predicted to have the property of not causing chromosomal changes (cannot change genetics).and does not trigger the growth of cancer cells (Hardjono, 2017).Classification of toxicity were classified based on Globally Harmonized System (GSH), which stated that toxicity was divided into six classes based on the LD50 value.Classes 1 -3 had a high level of toxicity, making them extremely hazardous , while grades 4 -6 had low toxicity, they were nonetheless potentially hazardous (Umboro et al., 2020).Thymidine compounds also showed inactive results on the parameters of hepatotoxicity, cytotoxicity and immunotoxicity.The thymidine compound showed good toxicity class results, namely class 6 (non-toxic), so based on the table it can be concluded that the compound that showed the best results was thymidine compound.

Analysis pharmacokinetic
Prediction of pharmacokinetic properties (Table 4) is carried out to see how effective a drug compound is in the body when consumed.Parameters observed included ADME (absorption, distribution, metabolism and excretion).
The results of the gastrointestinal (GI) absorption parameters of thymidine compounds showed high absorption.High absorption of a drug in the intestine can mean that the drug is good and can be distributed throughout the body (Lohidashan et al., 2018).Bioavailability score is a computational formulation to see the permeability of Caco-2 cells in the probability of a compound that can be well absorbed in the human body if it has a value of 0.55 and if it is 0.17 then it cannot be absorbed properly by the body (Martin, 2005).Based on the results of the bioavailability score of thymidine compounds obtained a value of 0.55.This shows that thymidine compounds can be well absorbed by the body in Caco-2 cells.The bioavailability radar can be seen in (Figure 5), the pink color represents the optimal range of the compound as an oral drug based on 6 parameters (lipophilicity: the ability of a chemical compound to dissolve into fat, size: 150-500 g/mol, polarity: TPSA value in the range of 20-130Å), solubility: logS value < 6, flexibility: rotatable bond < 9 and saturation: carbon fraction on sp3 hybridization < 0.25) (Daina et al., 2017).Based on the picture, it can be seen that the thymidine compound is in the pink area so it can be predicted that the compound has good oral bioavailability.Other pharmacokinetic parameters showed that the thymidine compound was predicted to have a high gastrointestinal (GI) absorption with PGPwhich indicates that once the compound enters the cell, it will not be pumped out again (Chelliah, 2008).This data is also reinforced by the results of the drug-likeness prediction (table) which shows that the thymidine compound complies with Lipinski's rule (Ro5) with 0 errors (Table 4).
To determine whether or not these compounds can be used as drug candidates / to determine the activity of a compound, it is not enough just to look at the binding affinity value of the molecular docking alone, but also a compound that has low toxicity, is easily absorbed by the body, has good permeability and has good properties.Good oral bioavailability.Based on data that has been obtained from several in silico tests, compounds that show the best results and have the potential to be developed as drug candidates are found in thymidine compounds, these compounds show stable ligand-receptor binding affinity values, good amino acid interactions, low toxicity values, It has good oral permeability and bioavailability.

CONCLUSION
The conclusion from the results of this study was that thymidine compound has ability for antiviral, highest binding score with DENV-3, and safe in terms of pharmacokinetics and toxicity.Thus this study recommends to conduct further in vitro and clinical studies of the thymidine compound present in Allium tuberosum to confirm its potential as a DENV-3 drug candidate.

Figure 2 .
Figure 2. The blue band represents the E protein DENV-3, while the green circle represents the active site of the E protein Figure 4. 2D visualization of ligand-receptor interactions (hydrogen bond: blue line; steric bond: red line; Red circle: same amino acid residue as native ligand)

Table 1 .
Validation of the docking method compounds in Allium tuberosum (Table2), the lowest RS value was found in thymidine compounds with a value of -81.1245 kcal/mol, meaning that thymidine compounds had the highest activity compared to other compounds.

Table 3 .
Toxicity prediction of thymidine compound, computed by ProTox-II