Inhibitory effects of selected isoquinoline alkaloids against main protease (Mpro) of SARS-CoV-2, in silico study.

The COVID-19 pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has become a global threat. Despite the production of various vaccines and different treatments, finding natural compounds to control COVID-19 is still a challenging task. Isoquinoline alkaloids are naturally occurring compounds known to have some potential antiviral activity. In this study, ten abundant isoquinoline alkaloids with antiviral activity were selected to analyze the preventive effect on COVID-19. A scrutinized evaluation based on Lipinski's rule showed that one out of ten compounds was toxic. Based on molecular docking analysis using Autodock software one of the best molecules with maximum negative binding energy was selected for further analysis. The Gromacs simulation analysis revealed that Coptisine has more action against active site Mpro of COVID-19. Overall, to make a rational design of various preventive analogues that inhibit the COVID-19, associated in vitro and in vivo analyses are needed to confirm this claim.

Introduction

COVID-19 has become a global pandemic to the health care system of almost every country in the world. In late December 2019, a new coronavirus was identified, originally named the 2019 new Corona Virus (nCoV 2019), which developed during the spread of the disease in Wuhan, Hubei Province, China (Wu et al. 2020; Narkhede et al. 2020; Kumar et al. 2020). In China, an abrupt prevalence was declared by The Emergency Committee of the World Health Organization (WHO) in late 2020 which was then recognized internationally as a public health emergency (Rodríguez-Morales et al. 2019; Chen et al. 2020). Afterward, the World Health Organization (WHO) has converted the appellation to coronavirus illness (COVID-19), in February 2020 (Khaerunnisa et al. 2020; Velavan and Meyer 2020; Cao 2020). According to the WHO, millions of people have been infected with the coronavirus, and this number is increasing. Lately, the complete number of the patient about around the world was reported to be 235,190,113 confirmed cases with more than 4,808,189 deaths (https://www.worldometers.info/coronavirus/).

One of the identified drug targets among the coronaviruses is main protease (Mpro), also known as 3CLpro. Amino acid sequences showed that there are many similarities between the main protease of COVID 19 with other SARS-CoV families and the majority of the sequences are conserved. This enzyme plays an important role in the processing of polyproteins translated from viral RNA and it is essential for the efficiency of the virus (Garg and Roy 2020; Alamri et al. 2020; Cao et al. 2020). Inhibition of this enzyme activity would prevent the proliferation of the virus. Despite the production of various vaccines and different treatments, finding natural compounds to control COVID-19 is still a challenging task. (Uddin et al. 2020; Cortegiani et al. 2020). Numerous inhibitors such as dipyridamole, candesartan cilexitil, hydroxychloroquine, chloroquine, disulfiram, atazanavir, indinavir, sulfacetamide, cimetidine, maribavir, and candesartan have been introduced to control SARS-CoV-2 Mpro (Li et al. 2020).

Isoquinoline alkaloids are important metabolites that contain a variety of biological and medicinal activities (Diamond and Desgagné-Penix 2016). Isoquinoline alkaloids are found in the plants (families of Papaveraceae, Fumariaceae, Menispermaceae, Annonaceae, and Rutaceae). These compounds are generally composed of the amino acids of PHE and TYR. Isoquinoline ring is an important feature in all of them (Bentley 2001). Different therapeutic approaches such as antiviral activity, antibacterial, anti-convulsant, anti-tussive, etc., are of special importance (Kukula-Koch and Widelski 2017). Previous studies have tested the effect of selected alkaloids against Mpro of COVID-19. The compounds such as Thalimonine, Lycorine, Hemanthamine, Berberine, Hippeastrine, Hirsutine, Fangchinolone, Tetrandrine, Cepharanthine, Skimmianine, and Emetine had the potential to inhibit the Mpro (Garg and Roy 2020). Therefore, in this study, ten isoquinoline alkaloid compounds were selected based on their antiviral activity reported in the literature. Chelidonine is found in Chelidonium majus and possesses diverse biological activities. Its antiviral activity was reported against immunodeficiency virus type 1 (HIV-1) (Da et al. 2015). Pscychotrine and Cephaeline both can be isolated from the roots of Psychotria nervosa and are being tested for potential activity against HIV-1 reverse transcriptase (Chinsembu 2019). Fumaricine is derived from Fumaria officinalis and act as an anti-virus (Dey et al. 2020). Galanthamine is an isoquinoline alkaloid that is extracted from Leucojum aestivum. It is a potent inhibitor of Herpes simplex type 1 viruses (Georgiev et al. 2012). Glaucine is found in the members of Papaveraceae family and shows various medical importance. Various studies reported its potential antivirus (anti-HIV) (Modarresi et al. 2020). Boldine was collected from Peumus boldus and has been investigated for the anti-influenza A virus (Zhao et al. 2006). Hydrastine and Coptisine from Ranunculaceae have been found as potential alkaloids against Hepatitis B virus (HBV) is the causative agent of B-type hepatitis in humans (Orhan et al. 2007). This study aimed to investigate the inhibitory nature of isoquinoline alkaloids on Mpro of COVID-19. Therefore, in silico parameters such as toxicity, Lipinski's rule, molecular docking, and molecular dynamics simulation were performed.

Methodology

Protease and ligands download

At first, the isoquinoline alkaloid molecules were fetched from the literature. The 3-dimensional (3D) structure of the compounds was download from the PubChem chemical database (http://pubchem.ncbi.nlm.nih.gov) (Kim et al. 2016) (Table 1). A crystallographic structure of COVID-19 main protease (PDB ID 6LU7) (Jin et al. 2020) with the resolution of 2.16 Å was obtained from a protein data bank (http://www.rcsb.org).

Table 1

Compound structure and PubChem CID of Isoquinoline alkaloids

S. no	Ligands	Compound structure	PubChem CID
1	Chelidonine		197,810
2	Psychotrin		3,496,498
3	Cephaeline		442,195
4	Fumaricin		609,998
5	Galanthamine		9651
6	Glaucine		16,754
7	Boldine		10,154
8	Drotaverine		1,712,095
9	Coptisine		72,322
10	Hydrastine		197,835

The toxicity of ligands

The study of the toxicity of compounds is an important feature in the selection of a drug molecule. In this study, the toxicity of isoquinoline alkaloids (Hepatotoxicity, Mutagenicity, Carcinogenicity, Cytotoxicity, and Immunotoxicity) and their toxicity class were predicted by the ProTox-II server (tox.charite.de/protox_II/) (Banerjee et al. 2018) and Toxtree 2.5.4 tool (Mombelli et al. 2016), respectively.

Evaluation of Lipinski's parameters

The potential effective of isoquinoline alkaloids compounds was evaluated using the Lipinski parameter to inhibit the activity of COVID-19 main protease. Items such as Molecular weight (Mw), high lipophilicity, hydrogen bonds donor, hydrogen bonds acceptor, and Molar refractivity were considered for the compounds (Walters 2012; Daina et al. 2017a). The SwissADME tools (http://www.swissadme.ch/index.php) (Sadeghi and Zarei 2020; Behbahani et al. 2021) and PubChem (http://pubchem.ncbi.nlm.nih.gov) were used to find Lipinski’s rule (Daina et al. 2017b; Sadeghi et al. 2021a).

Molecular docking

Autodock (version 4.2) (Sadeghi et al. 2022, 2021b) was used to perform molecular binding of each compound with 6LU7. Before molecular docking, the receptor and all compounds were optimized by Chimera software v1.7 (Sadeghi et al. 2021a; Kiffer-Moreira et al. 2014). Water molecules were removed, polar hydrogens were added and charged by Kollman charges. Then the grid box parameters were set to the appropriate scale (70 × 70 × 70 & spacing; 0.383). The 10 conformations were considered for each compound, and the best conformation (lower binding energy) was selected. Finally, Discovery Studio Visualizer (v16.2.0.16349) (Sadeghi and Zarei 2020; Heh et al. 2013) was used to show the two-dimensional image of the ligand-receptor complex.

Molecular Dynamics simulation (MDs)

Simulation of Mpro-ligand was studied using MDs for a time of 10 ns using GROMACS program v4.5.5. (Joshi et al. 2020). Gromacs 9643a1 was performed as the force field for the simulation of the Mpro-ligand complex. The Mpro-ligand complex was solvated in SPC/E Water Models (Mark and Nilsson 2001). After adding the water, the complex was neutralized by adding three ions Na+. The covalent contacts between the atoms were limited by short energy minimization. Then the system was equilibrated in two stages. The first stage contained a fixed number of particles, volume, and temperature (NVT). The second stage included a constant number of particles, pressure, and temperature (NPT). The simulation was performed at 300 K and 1 bar pressure for 10 ns. Covalent bonds and electrostatic interactions were adjusted by Linear Constraint Solver (LINCS) and Mesh Ewald (PME) methods, respectively. Molecular dynamics simulation protocol between receptor-ligand complex was considered by previous reports (Thirumal et al. 2017). Finally, the stability of the system was calculated by analysis the root mean square deviations (RMSD), an accessible radius of gyration (Rg), and root mean square fluctuations (RMSF).

Results and discussion

Toxicity analysis

All selected ligands were tested for toxicity (Table 2). One of the important medicinal properties of a compound is its non-toxicity. Therefore, the comparison of toxicity helps to remove toxic ligands. The toxicity of the compounds is divided into six different classes. The I class is highly toxic, the II and III classes are toxic, the IV class is low toxic and the V and VI classes are safe (Zhu et al. 2017). The results of analysis showed that among ten selected compounds, one compound had hepatotoxicity and immunotoxicity. So, this compound was removed and the remaining nine compounds were examined for Lipinski parameters.

Table 2

Evaluation of toxicity, toxicity class, and toxicity status of ligands

Ligands	Hepatotoxicity	Carcinogenicity	Immunotoxicity	Mutagenicity	Toxicity Class	Status
Chelidonine	−	−	−	−	V	N.T
Psychotrin	+	−	+	−	III	T
Cephaeline	−	−	−	−	VI	N.T
Fumaricin	−	−	−	−	V	N.T
Galanthamine	−	−	−	−	V	N.T
Glaucine	−	−	−	−	V	N.T
Boldine	−	−	−	−	V	N.T
Drotaverine	−	−	−	−	VI	N.T
Coptisine	−	−	−	−	VI	N.T
Hydrastine	−	−	−	−	V	N.T

A positive sign (+) indicates toxicity and a negative sign (−) indicates no toxicity

N.T; no toxic, T; toxic

Lipinski’s analysis

The ten selected compounds were compared according to Lipinski parameters (Table 3). These parameters remove ligands that may not be drug candidates. In addition to all of Lipinski's parameters of 5, the Veber filter was considered as another parameter for comparing of isoquinoline alkaloids. The eligibility of most parameters does not guarantee a specific substance as a drug. It only gives information about the drug-likeliness of the drug and only helps to eliminate weak ligands in the preclinical stage. The results of Lipinski parameter showed that all selected isoquinoline alkaloids follow this rule, and therefore 9 selected ligands were considered as drug candidates. The interaction was performed between Mpro ligand complexes.

Table 3

Lipinski and Veber filter analysis of ligands

Compounds	Hydrogen bond donors (≤ 5)	Hydrogen bond acceptors (≤ 10)	Molecular mass (< 500)	Logp (< 5)	Molar Refractivity (35–150)	Veber filter (30–80)
Chelidonine	1	6	353.4	2.2	96.12	60.39
Coptisine	0	4	320.3	3.5	87.95	40.80
Cephaeline	2	6	466.6	4.41	142.58	63.19
Fumaricin	1	6	369.4	2.28	102.92	60.39
Galanthamine	1	4	287.35	1.8	84.05	41.93
Glaucine	0	5	355.4	3.4	104.94	40.16
Boldine	2	5	327.4	2.7	96.00	62.16
Drotaverine	1	5	397.5	4.5	121.38	48.95
Hydrastine	0	7	383.4	2.7	103.38	66.46

Molecular docking is one of the important factors in drug design. In this method, the interaction between the receptor-ligand complexes is predicted, and the docking energy is investigated. The lowest docking energy creates the strongest binding (Sadeghi et al. 2021b). Natural products like coumarins, flavonoids, terpenoids, and diarylheptanoids are potent inhibitors of the Mpro (Paraiso et al. 2020). In silico and in vitro studies have found that catechin, quercetin, and gallocatechin are potent inhibitors of the Mpro (Mahmud et al. 2021). Mpro duplication was inhibited by isoquinoline alkaloids such as cepharanthine, berberine, fangchinoline, berbamine, tetrandrine, coptisine, palmatine, and jatrorrhizine (Chakravarti et al. 2021). Another in silico study showed that berberine, an isoquinoline alkaloid found in the leaves of Strychnos usambarensis (Gyebi et al. 2021), cryptoquindoline, and cryptospirolepine, two alkaloids found in Cryptolepis sanguinolenta (Borquaye et al. 2020), had high binding affinity for Mpro. The same study displayed that Hydroxyusambarensine revealed the strongest interactions with Mpro of SARS-CoV-2, and Amarogentin exhibited the highest binding affinity and selectivity for Mpro of SARS-CoV (Kar et al. 2020).

All the nine selected ligands were placed in Mpro active site. According to previous reports, the active site of the Mpro includes the amino acids of GLY143, SER144, CYS145, THR26, HIS163, THR27, and HIS172 (Garg and Roy 2020; Mpiana et al. 2020; Enmozhi et al. 2020). Docking scores were calculated for all compounds and compared with the standard inhibitor (N3). The binding energy of the N3-Mpro complex was − 8.17 kcal/mol. The binding energy of the ligands was ranged from − 4.9 to − 9.15 kcal/mol (Fig. 1). The results showed that all the compounds had a high potential for binding to Mpro. But Coptisine-Mpro complex had lower binding energy and stronger bond than N3-Mpro complex.

Fig. 1

The chart related of binding energy (-kcal/mol) of Isoquinoline alkaloids and N3 (Standard inhibitor) with Mpro

In the binding method, N3-Mpro complex via three conventional hydrogen bonds (THR26, CYS145, and SER144) as well as eleven van der Waals interactions. However, the Coptisine-Mpro complex had better binding energy via four conventional hydrogen bonds (GLY143, SER144, CYS145, and HIS163) and twelve van der Waals interactions (Fig. 2A, B, C, D).

Fig. 2

2D and 3D interactions of Mpro and two inhibitors: A 2D interaction of Coptisine-Mpro complex. B 3D interaction of Coptisine-Mpro complex. C 2D interaction of N3-Mpro complex. D 3D interaction of N3-Mpro complex

Molecular dynamics simulation analysis

Molecular dynamics is used to study ligand-receptor complexes overtime at the atomic level. various factors such as RMSF, RMSD, and Rg help in understanding the binding template (Sneha and Doss 2016). Figure 3 depicts the RMSF factor of the N3-Mpro complex and Coptisine-Mpro complex. RMSF studies the flexibility among the residues in the presence of Coptisine and N3. From the RMSF plot, it was perceived that the Coptisine-Mpro complex (fluctuation total = 31.7103) showed a lower fluctuation level than the N3-Mpro complex (fluctuation total = 33.8953) indicates the restricted movements during the simulation.

Fig. 3

Root mean square fluctuation (RMSF). Orange color indicates Coptisine-Mpro complex; Blue color indicates N3-Mpro complex

Examining RMSD factors provides accurate structural information in understanding the structural stability of any complex. Therefore the RMSD analysis for Mpro was performed in the presence of Coptisine ligand as well as N3 inhibitor. According to the results, the equilibrium of both complexes is established up to 5 ns, but after 5 ns the Coptisine-Mpro complex shows less deviation. N3-Mpro complex achieves equilibrium at 0.19 nm at 10 ns whereas the Coptisine-Mpro complex achieves equilibrium at 0.16 nm at 10 ns (Fig. 4). The lower RMSD factor of the Coptisine-Mpro complex indicates higher stability of the Coptisine-Mpro complex structure than that of the N3-Mpro complex.

Fig. 4

Root mean square deviations (RMSD) levels for Coptisine-Mpro complex and N3-Mpro complex during 10 ns molecular dynamics simulation. Blue color indicates Coptisine-Mpro complex; Green color indicates N3-Mpro complex

To observation the total compactness of resistance Mpro in the presence of Coptisine and N3, we performed the Rg level of Coptisine-Mpro complex and N3-Mpro complex (Fig. 5). From the plot, it is obvious that the Coptisine-Mpro complex showed a lesser Rg level than the N3-Mpro complex, demonstrated the low conformational changes during the simulation. This lesser the Rg clarify the complex had more compactness and contrariwise.

Fig. 5

radius of gyration (Rg). Green color indicates Coptisine-Mpro complex; Blue color indicates N3-Mpro complex

Conclusion

In this study, the aim was to find effective inhibitors among isoquinoline alkaloids against the main protease of COVID 19. Four filtering steps were considered for this purpose. The toxicity of ten ligands was evaluated and nine compounds entered the Lipinski stage. Molecular docking was performed for nine compounds. Among the isoquinoline alkaloids compounds, the Coptisine had the best binding energy compared to the standard inhibitor with a binding energy of − 9.15 kcal/mol. Therefore, molecular dynamics simulations were considered for the Coptisine-Mpro complex and N3-Mpro complex. RMSD, RMSF, and Rg values were compared for both complexes. It can be concluded that the Coptisine as a natural inhibitor can have a potential effect on the Mpro of COVID-19 in silico. The tests in vitro and in vivo are essential to ensure the inhibitory effect of the desired compound.