IN SILICO ANALYSIS OF SECONDARY METABOLITES OF Clerodendrum Inerme AS A POTENTIAL ANTIDIABETES COMPOUNDS

Clerodendrum inerme can potentially alleviate diabetes, but little is known about its molecular mechanisms. This study aimed to investigate the chemical compound of C. inerme and its molecular mechanism to treat diabetes. The KNApSAcK was used to find secondary metabolite of C. inerme . A screening was done to find compounds by estimating Absorption, Distribution, Metabolism, and Excretion (ADME) on the SwissADME. The SwissTargetPrediction tool connects predictions of target proteins from compounds that pass screening to various probable proteins and utilizing the StringDB to show the network between target proteins and associated diseases. After finding the target protein, continue docking the chemical compound to the target protein using PyRx with AutoDock 4.2.6. The result from StringDB found four chemical compounds ((Z)-3-Hexenyl beta-D-glucopyranoside, Rhodioloside, Sammangaoside B, Clerodermic acid) that can connected to 4 target proteins (DPP4, IL1B, PPARA, PPARG). According to the docking results, clerodermic acid has good protein binding properties with DPP4, IL1B, PPARA, PPARG, rammangaoside B with PPARG, and rhodioloside with DPP4. C. inerme contains clerodermic acid, rammangaoside B, and rhodioloside compounds, which can potentially treat diabetes mellitus.


Background
Diabetes mellitus is a collective term for heterogeneous metabolic disorders whose main finding is chronic hyperglycemia.The cause is insulin secretion disorder, insulin disorder effect, or usually both (Petersmann et al., 2019).Diabetes mellitus is classified into three types.Type 1 diabetes causes cell damage, resulting in the body's inability to produce insulin.Insulin resistance, a condition in which cells fail to respond to insulin properly, is the starting point of type 2 diabetes (Tanase et al., 2020).experienced by pregnant women is due to a decrease in the body's ability to produce insulin to control blood sugar levels during pregnancy (Setiabudy, Nafriadi and Instiaty, 2016).The goal of diabetes mellitus treatment is to achieve normal levels of insulin in the plasma (Ferguson and Finck, 2021).According to IDF data, by 2021, 537 million adults aged 20-79 years are diagnosed with diabetes.The number is expected to rise to 643 million by 2030 and 783 by 2045.In Southeast Asia, 90 million adults have diabetes, which causes 747,000 deaths (Webber, 2021).New therapies must be invented to meet the needs of health care, including promotion, prevention, treatment, and rehabilitation, so that the prevalence of diabetes mellitus can be decreased.Dozens or even hundreds of new drugs are released to the market every year after going through a time-consuming and expensive development process (Hairunnisa, 2019).Clerodendrum inerme, commonly known as gambir laut in Indonesia, belongs to the Verbenaceae family.It is commonly found in Australia, Asia, Malaysia, and the Pacific Islands.C. inerme is traditionally used to treat malaria.It is also used as a thermal suppressant, uterine stimulant, pest control agent, and antiseptic (Kar et al., 2019).Although it is not clear that this plant has antidiabetic effects, it is a good plant to study the contents of secondary metabolite compounds with antidiabetic effects.This research aims to find new drugs with antidiabetic effects through network pharmacology and molecular docking.The method is used as an early stage of research before further in-vivo research.

Tools
This study was conducted using several online databases and software.

Secondary Metabolite of C. inerme Identification and Network Pharmacology Analysis
The secondary metabolite of C. inerme was obtained from KNApSAcK, and PubChem was used to search for the canonical SMILE compounds (Kim, 2021).Screening of compounds using the SwissADME website that will predict the bioavailability of the compound using the BOILED-egg method.Furthermore, SwissTargetPrediction was used to predict proteins that can interact with secondary metabolite compounds (Lena et al., 2023).Search for protein targets that interact with diabetes was carried out using the GeneCards (Stelzer et al., 2016).followed by looking for protein intersections predicted to have ties to compounds from the plant using the Venny database (Oliver, 2015).Furthermore, the intersection results are entered into the StringDB for network pharmacology analysis (Szklarczyk et al., 2021).

Molecular Docking Analysis
Docking molecules using 3D files obtained from PubChem and prepared using Avogadro by MMF94s Method.Separation of 3D files between the target protein of diabetes mellitus and its ligand was done using BIOVIA Discovery Studio.Molecular docking was done using PyRx 0.8 with autodock 4. The results of the docking are then visualized using the Proteins.Plus webserver.

Identification, Bioavailabity Prediction, and Network Pharmacology Analysis of Secondary Metabolite of C. Inerme
The secondary metabolite of C. inerme was obtained using the KNApSAcK database.There were found 24 metabolite compounds contained in the C. inerme.SMILES canonical code was searched using PubChem, but six of these compounds were not found in its SMILES code, so they were not included in the study (Table 1).Then, the next stage is carried out, namely, selecting compounds based on ADME.The term used is Lipinski's Rules of Five (RoF), which says that in compounds which have a molecular weight lower than 500 Da, the number of hydrogen bond donors is less than 5, the number of hydrogen bond acceptors is less than 10, and  log  is lower than 5 have high bioavailability (Nogara et al., 2015).From Lipinski's RoF, it was obtained four compounds: Mol 6, Mol 10, Mol 12, and Mol 21.Besides that, the compounds were also examined using the Brain Or Intestinal EstimateD (BOILED-Egg) method.For this purpose, BOILED-Egg is proposed as an accurate prediction model that calculates the lipophilicity and polarity of small molecules (Daina and Zoete, 2016).
Proceedings of International Pharmacy Ulul Albab Conference and Seminar (PLANAR), 3, 57-65 60 From these results, one compound (Mol 21) penetrated the blood-brain barrier, marked in the yellow section.The other three compounds (Mol 6, Mol 10, and Mol 12) are in the white part, meaning these compounds cannot penetrate the blood-brain barrier but can be absorbed in the digestive tract.

Figure 1. Bioavailability prediction with BOILED-Egg Method
The four compounds that pass ADME will be checked to see whether they can bind to proteins using SwissTargetPrediction.The prediction results are a percentage probability of the target protein binding.After knowing the target protein of the plant compound, continue comparing it with the Diabetes Mellitus-related protein obtained from GeneCards.178 proteins were predicted to interact with secondary metabolite compounds, and there are found 15,390 diabetes mellitusrelated proteins.Intersection with Venny found that only 159 proteins from SwissTargetPrediction were related to Diabetes Mellitus.

Molecular Docking Analysis
Molecular docking was carried out between the four secondary metabolite that pass Lipinski's RoF and BOILED-Egg, with the proteins DPP4 (PDB ID = 6EOR), IL1B (PDB ID = 6Y8I), PPARA (PDB ID = 3KDT), and PPARG ((PDB ID = 8HUP).More complete docking results can be seen in table 4. The docking results obtained only a few have the highest potential as anti-diabetics.A compound can be said to be good if the energy binding value and the inhibition constant value were low.The lower the value of the energy binding and inhibition constant, the better the compound binds to protein (Muchlisin et al., 2022).Docking result that have good potential only Mol 10 to DPP4 (-8.16 kcal/mol, 1.04 µM), Mol 12 to PPARG (-6.98 kcal/mol, 7.6 µM), and Mol 21 to DPP4 (-7.79 kcal/mol, 1.94 µM), IL1B (-7.63 kcal/mol, 2.56 µM), PPARA (-9.07 kcal/mol, 225.97 ηM), and PPARG (8.09 kcal/mol, 1.17 µM) which has high potential as an anti-diabetic drug candidate.The bonds formed between the compound and the target protein are in the form of hydrogen and hydrophilic bonds which can be seen in table 5.In the docking results of Sammangaoside B to PPARA, no hydrogen or hydrophilic bonds occurred.

Conclusion
The compounds Rhodioloside, Sammangaoside B, Clerodermic Acid have potential as new antidiabetic drugs by binding to the proteins DPP4, IL1B, PPARA, PPARG.

Figure 3 .
Figure 3. Network Pharmacology prediction results using String-db.The red color shows the target protein associated with Diabetes Mellitus.

Figure 4 .
Figure 4. Network pharmacology of target prediction protein for Diabetes Mellitus

Table 1 .
List of metabolite compounds contained in C. inerme taken from the KNApSAcK website.