Structure, activity, and drug-likeness of pure compounds of Sumatran lichen (Stereocaulon halei) for the targeted ACE2 protein in COVID-19 disease

Sumatran lichen has the potential as antiviral, pure isolates that have been isolated and developed as prospective compounds for COVID-19 treatment. Computational methods were used to accelerate the discovery and screening of potential new compounds. The molecular structures of the isolated compounds such as Lobarin, Atranorin, Methyl 2,4-dihydroxy-3,6-dimethylbenzoate, Methyl 3‐formyl‐2,4‐dihydroxy‐6‐methylbenzoate, Ethyl 3-formyl-2,4-dihydroxy-6-methylbenzoate, and Lobaric acid were drawn, then their activities were analyzed, processed by docking with ACE2 protein, and tested for Druglikeness. The activities and druglikeness were determined in the Swiss ADME program, while the ACE2 docking was processed by Blind Docking in Arguslab, AutoDock Vina, Open Babel, and Discovery Studio Visualizer programs. All compounds bound to the ACE2 protein, as apparent from the number of hydrogen bonds between the two. The Gibbs free energy was in the range of -5.6 to -7.0, and the best one was obtained from atranorin. As for lobarin, this compound was found to be non-drug-like.


Procedures ADME and toxicity
The 2D structure was drawn in the Marvin Sketch program, and the compound was analyzed according to the name listed in the International Union of Pure and Applied Chemistry (IUPAC). The protonation was adjusted at pH 7.4, which is the normal condition of the human body. Then, the chemical structure was analyzed by SwissADME (Absorption, Distribution, Metabolism, and Excretion). As for the toxicity, it was predicted using the structural alert feature provided in QSAR Toolbox.

Ligands and proteins
The materials used were the 3D macromolecular structures of Angiotensin-converting enzyme 2 (ACE2) (PDB ID: 6M17), obtained from the online database of Protein Data Bank (www.pdb.org), and compounds isolated from Stereocaulon halei.

Ligand preparation
The ligands used were the isolated compounds of Stereocaulon halei. Their 3D structures were built and processed by the semi-empirical method (PM3) geometry optimization using the Arguslab program. The optimized structures were saved in *.xyz and then converted to *.pdb using the OpenBabel program

Receptor preparation
The ACE2 macromolecules (GDP: 6M17) were downloaded from the protein data bank (GDP) website, www. rcsb.org. The structure of this protein was displayed in the Discovery Studio Visualizer v.4.5 program package, separated from the water molecules and their natural ligands, then stored in *.pdb and treated as a receptor.

Molecular simulation
Subsequently, the test compounds were tethered in the receptor using the blind docking method. This method was used because the test receptors were categorically new, still in the process of research and development, and, therefore, had no congenital ligands. Besides, the active side of ACE2 was unknown. The molecular docking was carried out using the AutoDock Vina program. The results of the analysis were free energy bonds, hydrogen bonds, and binding patterns with other amino acid residues at the test receptors.

RESULTS AND DISCUSSION Predictions of the chemical-physical properties
The predicted chemical-physical properties of the pure compounds of Stereocaulon halei were presented in Table 1. These properties included the predicted values of the molecular weight, molar refractivity, topology surface area (TPSA), log p, and log S. All compounds had molecular masses in the range of 196.20-474.50 g/mol. An orally administered drug can be easily absorbed if the molecular mass is less than 500 g/mol (Lipinski, 2004). Molar refractivity is a measure of the steric factor, i.e., the volume occupied by an atom or group of atoms. It is used as a measure of substituent steric influence in the Quantitative Structure-Activity Relationship (QSAR) equation (Patrick, 2013).
Topology Polar Surface Area (TPSA) is a 2D-QSAR descriptor that represents a relationship between ligands and specific targets (Prasanna and Doerksen, 2008). Molecules with TPSA values higher than 140 angstroms are inclined to permeate the cell membrane poorly (Pajouhesh and Lenz, 2005). The same case applies to lobarin.
The log p-value reflects the tendency of a fat-soluble compound. All compounds isolated from Stereocaulon halei had log p of <5, meaning that they can be easily absorbed through oral administration (Lipinski, 2004). Here, log p was the consensus value from various log p determination methods. Log S is the logarithm of water solubility. Based on the log S values, lobarin, atranorin, and lobaric acid were categorized as compounds with low water solubility. Meanwhile, methyl 2,4dihydroxy-3,6-dimethylbenzoate, methyl 3-formyl-2,4-dihydroxy-6-methylbenzoate, and ethyl 3formyl-2,4-dihydroxy-6-methylbenzoate were identified as water-soluble compounds.

Pharmacokinetic prediction, drug-likeness, and medicinal chemistry
The pharmacokinetic model represents the relationship between chemical coordination in tissues, chemical calculations, and ADME processes (El-Masri, 2013). The predicted pharmacokinetic profiles of the compounds observed are summarized in Table 2. After oral absorption, the drug permeates the gastrointestinal mucous membrane then enters the blood circulation (Hirtz, 1985). Lobarin had low absorption in the low gastrointestinal membrane, Lobarin had low absorption in the low gastrointestinal membrane, in which this compound is found at high levels in the blood. Also, lobarin has four OH groups that can decrease gastrointestinal absorption.
Methyl 2,4-dihydroxy-3,6-dimethylbenzoate was able to cross the blood-brain barrier (BBB) due to the presence of a methyl group in the para position. Compounds permeating BBB by diffusion through lipids have molecular weights of <400 Da and hydrogen bonds of <8 (Pardridge, 2012). P-glycoprotein substrates are one of the drug-carrying proteins that affect drug distribution. This process affects plasma P -glycoprotein substrates as a transmembrane pump, which pumps the substrate from within to outside the cell (Finch and Pillans, 2014).
Lobaric acid had the highest log Kp, meaning that it has the slowest skin penetration rate among the other compounds. On the contrary, methyl 3-formyl-2,4-dihydroxy-6-methylbenzoate had the lowest log Kp value or, in other terms, the fastest skin penetration rate. Log Kp is the logarithmic coefficient of skin permeability (Kp) expressed in cm/s. In this study, it was obtained using the quantitative structure-activity relationship (QSAR) method by modeling the influencing factors of skin permeability, including molecular mass, melting point, and the log value of the octanol-water partition coefficient (Chang et al., 2012). Drug-likeness is the balance of molecular and structural properties that determine whether or not a drug molecule is similar to a currently or previously used drug. This study used the drug-likeness analysis proposed by Egan (Pharmacia) and Muegge and bioavailability prediction. The results are presented in Table 3. The Egan's drug-likeness estimation proved that Lobarin did not meet the requirements of drug-likeness because its TPSA was above 131.6. Meanwhile, the Muegge's druglikeness database showed that there were three compounds, including Lobarin, Methyl 2,4-dihydroxy-3,6-dimethylbenzoate, and lobaric acid, that did not meet the requirements of drug-likesness because Lobarin and Lobaric Acid have log p values were greater than 5. Methyl 2,4-dihydroxy-3,6dimethylbenzoate has molecular weights were less than 200 g/mol. Bayern also uses Muegge's druglikeness method (Daina et al., 2017).
The bioavailability values of the compounds were 0.55 and 0.56. Compounds No. 2,3,4, and 5 had the bioavailability value of 0.55 and not negative charge (at pH 6); hence, the requirements of good medicine were met. Compounds No. 1 and 6, i.e., lobarin and lobaric acid, had the bioavailability value of 0.56, negative charge of -1 or -2 (at pH 6), and PSA between >75 and <150 Å 2 (Martin, 2005).
No compounds were categorized into Pan-Assay Interference Compounds (PAINS). PAINS are compounds that exhibit specific targets and often give false-positive results (hit identification) in High Throughput Screening (HTS) screening (Koptelov et al., 2018). The Brenk filters, containing a database of 105 fragments (Brenk et al., 2008), aims to find structural alerts of potential toxic effects of the compound observed. Structural alerts are chemical structures that indicate the toxic nature of a compound, in which part of the structure that has mutagenic and carcinogenic potentials is identified. The aldehydic group can lead to the formation of a Schiff base with a primary amine. This is considered potentially genotoxic, as demonstrated in vivo by the ability to react with nucleobases without metabolic activation, interbase cross-links, and adducts formation (Speit et al., 2007) According to an in vitro test at the HGPRT locus of V79 Chinese hamster cells, phenol is reported to have mutagenic properties in the liver (Paschin and Bahitova, 1982). Based on the Brenk's filters, atranorin contained structural alerts of aldehyde and phenol structural alerts; methyl 3-formyl-2,4-dihydroxy-6-methylbenzoate and ethyl 3-formyl-2,4-dihydroxy-6-methylbenzoate had structural alerts of aldehyde; and lobaric acid had structural alerts of phenol.
The degree to which a compound is difficult to synthesize ranges between 1 (very easy) to 10 (very difficult). Its determination takes into account the factors of complexity, such as macrocyclic, chiral center (Ertl and Schuffenhauer, 2009). Based on this degree, the most easily synthesized compound was compound No. 3, followed by 4,5,2,6, and then 1.

Docking
The docking simulation produced 20 docking poses with different binding energies that were mostly negative because of the presence of strong interaction. The Gibbs free energies identified in this study are presented in Table 4.
The best compound that can inhibit receptors by spontaneously forming bonding energy was selected according to the lowest Gibbs free energy. Based on the results of the study, Atranorin had the lowest Gibbs free energy (-7.0 kcal/mol) among the other test ligands. Hydrogen bonding is an interaction that occurs between two molecules, each acting as a donor and an acceptor. Hydrogen bonds are formed as hydrogen binds atoms like fluorine (F), nitrogen (N), and oxygen (O) (Głowacki et al., 2013).
Root Mean Square Deviation (RMSD) was calculated to quantify the experimental value that evaluates whether or not the docking program can produce a pose that is close to the native conformation. A real-native conformation is achieved if RMSD relative to the native pose is ≤ 2.0 Å (Bell and Zhang, 2019). At this state, the docking poses produced are expected to have the lowest RMSD cut-off value ≤ 2.0 Å (Angstrom).
As visualized and calculated using the Discovery Studio program, the hydrogen bonding between the test ligand and amino acid residues at the receptor was indicated by a green line. The identified ligand-receptor interactions are presented in Table 5. Apart from the formation of hydrogen bonds between ligands with protein, visualization using the Discovery Studio 2016 also displayed any amino acid residues involved in hydrophobic interactions. The amino acid residues that interacted with the ligand are listed in Table 6.

CONCLUSIONS
Lobarin does not fulfill the requirements of the Egan and Muegge's drug-likeness. The compounds isolated from Stereocaulon halei are Atranorin, Lobarin, Methyl 2,4-dihydroxy-3,6dimethylbenzoate, Methyl 3-formyl-2,4-dihydroxy-6-methylbenzoate, Ethyl 3-formyl-2,4-dihydroxy-6-methylbenzoate, and Lobaric acid, with the Gibbs free energy of -7.0, -6.7, -5.9, -5.9, -5.6, and -5.6 kcal/mol, respectively. Based on the results of screening and visualization between the ligand and the protein, the general interactions that occur are in the form of hydrogen bonds and hydrophobic interactions. Further research is required so as to find out other parameters and the active site of the enzyme that a good inhibitor must have and to determine the stability of the ligand-protein bonds, especially in Atranorin.