The main protease (Mpro) is an indispensable enzyme responsible for viral replication and it is among the top characterized drug targets of coronaviruses 35. Upon host infection by SARS-CoV-2, replicase polyproteins 1a and 1ab are synthesized to produce various functional subunits essential for viral replication. This is accomplished by site specific hydrolysis of the polyproteins by two viral proteases, one of which is the Mpro 36. Inhibition of this enzyme has been recognized to be an important strategy for therapeutic intervention against the SARS-CoV-2. In this study, in silico techniques were employed to identify possible SARS-CoV-2 inhibitors among the compounds of the Artemisia annua, an antimalarial plant with reported activity against the SARS-CoVs.
The compounds of Artemisia annua showed varying levels of binding affinities for the SARS-CoV-2 main protease in the molecular docking analysis, the highest being -7.83 kcal/mol by 7-Methylkaempferol (Rhamnocitrin) followed by -7.81 kcal/mol by 3-Methylkaempferol (Isokaempferide) and -7.65 by Kaempferol. It is worth noting that all the ten top-scoring compounds are flavonoids, which mainly include Kaempferol and its derivatives and Quercetin derivatives. Kaempferol, a tetrahydroxyflavone with its four hydroxy groups at positions 3, 5, 7 and 4' is a plant-derived aglycone flavonoid commonly found in medicinal herbs, fruits, seeds and vegetables. This phytochemical has been demonstrated to possess several pharmacological actions, among which are, antioxidant, anti-inflammatory, anticancer and antiviral activities 37. Kaempferol has been found to display antiviral activity against several viruses including coronavirus 38,39. A recent study by Xia et al. identified Kaempferol and Quercetin along with Luteolin as the main active ingredients of the Chinese herbal combination (Amygdalus Communis Vas and Ephedra sinica Stapf) used in the treatment of COVID-19 40. They also reported that these active ingredients demonstrated good binding affinity for SARS-CoV-2 Mpro 40 as substantiated in this present study.
The reliability of the molecular docking result was determined by estimating the binding free energy through the Prime MM-GBSA module of Maestro. This approach is one of the most reliable methods of validating docking results as it helps to determine the stability of the receptor-ligand complex 41,42. This denotes that a favorable binding free energy correlate to a reliable output from molecular docking study. This post docking analysis showed that the ten top-scoring phytochemical constituents of Artemisia annua possess rich binding free energy values (-39.67 kcal/mol and -50.61 kcal/mol) towards SARS-CoV-2 main protease. Studies have shown that the higher the binding free energy, the more favorable and stable the ligand-bound protein 43,44. With this in mind, Astragalin showed the highest stability with the SARS-CoV-2 main protease, followed by Rhamnocitrin and Quercimeritrin. The ten selected compounds could however be said to form favorable complexes with the protein crystal structure of SARS-CoV-2 main protease. This can be further explained from the scatter plot on figure 5, as most of the points are close to the regression line, which is also an indication that the molecular docking result is comparable to a large extent with the binding free energy of the compounds.
The inhibitory potentials of Artemisia annua compounds against SARS-CoV-2 Mpro was further revealed by the interactions of Rhamnocitrin, Isokaempferide and Kaempferol with the active site amino acid residues of the enzyme (figures 3 and 4). HIS 41 and CYS 145 to which these compounds bind are known to play very crucial roles at the active site of the SARS coronavirus Mpro. Previous studies on the crystal structure of the SARS-CoV-2 main protease showed that the protein is composed of three domains: domain I comprising of residues 10−99, domain II comprising of residues 100−182 and domain III comprising of residues 198−303. The active site holds a HIS 41-CYS 145 catalytic dyad in a cleft between domains I and II, with CYS 145 acting as a nucleophile during the first step of the enzymatic process and HIS 41 acting as a base catalyst 45,46. Also of interest is the interaction of the compounds with GLU 166 which is essential for Mpro substrate-induced dimerization required for catalysis and ASN-142 which forms hydrogen bonding with GLU 166 to block the substrate-binding subsite entrance in the monomer. Mutation of GLU 166 is reported to significantly block the substrate-induced dimerization process, thereby preventing enzyme activation 47. Therefore, interaction of Artemisia annua Kaempferol and its 3 and 7 methyl derivatives with these amino acid residues, which are targets of most SARS-CoV-2 main protease inhibitors, further validates them as potential therapeutic agents against the virus.
The in silico pharmacophore modeling was employed to identify the structural features of the test compounds responsible for their affinity for the target protein. This is to further authenticate the inhibitory potentials of the compounds against SARS-CoV-2 main protease. Two aromatic rings and one hydrogen bond donor were identified as the structural features of Rhamnocitrin, Isokaempferide and Kaempferol responsible for their interactions with the enzyme. The involvement of aromatic rings, in addition to hydrogen bond formation, in the interaction of the compounds with the enzyme might have contributed to the higher binding affinity of the compounds compared with the standard ligand which depended on two hydrogen bond donors only. Aromatic rings are vital residues for molecular interactions and frequently exist in several protein–ligand and protein–protein interactions. Owing to their natural existence in amino acids residues like histidine, tryptophan, phenylalanine and tyrosine, they are considered to be very important for protein stability and molecular recognition processes. Furthermore, aromatic rings are frequently used in drug design due to their role in the improvement of binding affinity and specificity of drug-like molecules 48.
In spite of the Mpro inhibitory potential exhibited by the Artemisia annua compounds, the ADMET properties of the compounds is an essential factor that will determine their pharmacological activity against SARS-CoV-2. In silico ADMET prediction is a fast and low-cost approach to determine whether the compounds will be easily absorbable, well distributed to their target site of action, favorably metabolized and easily eliminated from the body without leaving toxic side effects 49. The Lipinski filter has been an effective method for screening potential drug candidates for oral drug-likeness based their molecular weights, hydrogen bond acceptors and donors and lipophilicity 50. Therefore, compounds like Rhamnocitrin, Isokaempferide, Kaempferol, Apigenin, Penduletin and Isorhamnetin with zero Lipinski violations are likely to be orally active and this is further validated by their oral bioavailability scores. The 0.55% bioavailability score signify that these compounds have about 55% probability of a minimum of 10% oral absorption in rat or human colon carcinoma absorptivity 51. The remaining four compounds with 2 Lipinski violations and bioavailability score of 0.17 are not likely to be orally active. This is also reflected in the GI absorption potential of the compounds which is high for Rhamnocitrin, Isokaempferide, Kaempferol, Apigenin, Penduletin and Isorhamnetin, but low for the remaining compounds. However, none of the compounds show BBB permeability.
Despite its oral bioavailability potential as explained above, Isorhamnetin as a Pgp substrate is likely to be restricted from entering its target site of action. Pgp is a member of the ATP-binding cassette transporters that plays a role of active efflux of foreign chemicals through cell membranes in order to protect the body from these chemicals, and it is a major factor contributing to drug resistance. Hence, the bioavailability and consequent therapeutic effectiveness of Isorhamnetin might be limited by this factor. Additionally, the cytochrome P450 inhibitory potentials shown by Rhamnocitrin, Isokaempferide, Kaempferol, Apigenin, Penduletin and Isorhamnetin is an indication that the compounds could cause drug-drug interactions. This is because more than 50% of drugs are metabolized by these CYP isoforms and their inhibition, which could hinder the metabolism of these drugs, is a major cause of pharmacokinetics-related drug-drug interactions 52.
It is important to note that the toxicity prediction of the compounds appear to be favorable (table 4). All the compounds belong to the oral toxicity class 5, and their LD50 is between 2,500 and 5000 mg/kg, which means they could be safely used within these dosage limits. Furthermore, none of the compounds is likely to be hepatotoxic, carcinogenic, mutagenic and cytotoxic, making them relatively safe as potential therapeutic agents against SARS-CoV-2.
In conclusion, out of 168 bioactive compounds of Artemisia annua screened for possible inhibitory activity against SARS-CoV-2 main protease, Rhamnocitrin exhibited the highest binding affinity followed by Isokaempferide and Kaempferol. The three compounds, like the standard ligand occupied the active site of Mpro, where they interacted with important amino acid residues like HIS 41, ASN 142, CYS 145 and GLU 166, among others. Two aromatic rings and one hydrogen bond donor are involved in the molecular interactions of the three test compounds with the enzyme. The compounds also possess favorable ADMET profile and none of these three compounds showed the tendency for hepatotoxicity, carcinogenicity, mutagenicity, cytotoxicity and immunotoxicity. Therefore, these Artemisia annua compounds could be considered for experimental studies and further development into drugs for the treatment of SARS-CoV-2.