Interaction of Drug Candidates with Various SARS-CoV-2 Receptors: An in Silico Study to Combat COVID-19.

The world is currently facing the COVID-19 pandemic caused by the SARS-CoV-2 virus. The pandemic is causing the death of people around the world, and public and social health measures to slow or prevent the spread of COVID-19 are being implemented with the involvement of all members of society. Research institutions are accelerating the discovery of vaccines and therapies for COVID-19. In this work, molecular docking was used to study (in silico) the interaction of 24 ligands, divided into four groups, with four SARS-CoV-2 receptors, Nsp9 replicase, main protease (Mpro), NSP15 endoribonuclease, and spike protein (S-protein) interacting with human ACE2. The results showed that the antimalarial drug Metaquine and anti-HIV antiretroviral Saquinavir interacted with all the studied receptors, indicating that they are potential candidates for multitarget drugs for COVID-19.

Introduction

The end of 2019 and the beginning of 2020 were marked by the pandemic caused by the severe acute respiratory syndrome coronavirus 2. The first cases of infection by a novel coronavirus (SARS-CoV-2) emerged in December 2019 and were related to exposure to the Huanan Seafood Wholesale Market in the city of Wuhan, Hubei, China.[1,2] The COVID-19 then spread worldwide, and by July 14, 2020, there were 12 964 809 confirmed cases and 570 288 deaths globally.[3] Like SARS-CoV, SARS-CoV-2 seems to use the ACE2 receptor to enter the target cells.[4−8] Alveolar epithelial type II cells are 83% of ACE2-expressing cells,[5] which may explain the higher damage of the infection caused to the lungs, where COVID-19 showed high expression among alveolar epithelial in immunostaining study.[9] There is also evidence that a furin-like cleavage site in the S-protein of SARS-CoV-2 may be implicated in viral invasion of the cells.[10]

In most severe cases of COVID-19, patients require endotracheal intubation and mechanical ventilation, as they may develop severe pneumonia, pulmonary edema, ARDS, sepsis shock, or multiple organ failure.[2−11] The disease affects most severely males more than females, and patients with chronic cardiovascular and cerebrovascular diseases, diabetes, obesity, smoking history, and other comorbidities, resulting in lower immunity, have the worst prognostics.[12−14]

Research institutions are accelerating the discovery of vaccines and therapies for COVID-19. In this work, molecular docking was used to study (in silico) the interaction of 24 ligands with four SARS-CoV-2 receptors. Crystallographic structures of SARS-CoV-2 receptors, Nsp9 replicase protein, main protease (Mpro), and NSP15 endoribonuclease were used. One of the receptors was the model built by homology modeling of the spike protein (S-protein) and the human ACE2 receptor, and obtained in a study by Smith and Smith, 2020.[15] The objective of this work was to study the interactions performed by the ligands on these receptors and to indicate drug candidates for COVID-19.

Materials and Methods

3D Structure of Ligands and Receptors

The 3D structures of the ligands were obtained from the PubChem Open Chemistry Database (https://pubchem.ncbi.nlm.nih.gov/) (see Table ). The ligands were divided into four groups. Group 1 includes six ligands that appear in the recent literature on COVID-19. Group 2 presents the three ligands that obtained the best results in the study by Smith and Smith, 2020.[15] Group 3 is made up of five anti-HIV antiretrovirals, and Group 04 is made up of ten ligands from the database of our research laboratory used in previous studies (Table ).

Table 2

Results of Molecular Docking with Vina

groups	ligand	PubChem CID	receptor 1 ΔGbind (kcal/mol)	receptor 2 ΔGbind (kcal/mol)	receptor 3 ΔGbind (kcal/mol)	receptor 4 ΔGbind (kcal/mol)	references
01	Hydroxychloroquine	3652	–5.0	–5.8	–6.1	–5.4	(19−21)
	Azithromycin	447043	–4.9	–5.8	–5.6	–6.7	(19,22−24)
	Mefloquine	40692	–6.8	–7.6	–7.3	–6.7	(25)
	Metaquine	10670321	–7.5	–8.1	–8.6	–7.9	(26,27)
	Nelfinavir	64143	–6.1	–7.6	–7.7	–8.0	(28,29)
	PL-69	25208163	–7.4	–7.2	–7.6	–6.8	(30)
02	Pemirolast	57697	–5.8	–6.5	–6.0	–7.4	(15)
	Benserazide	2327	–4.9	–5.9	–5.6	–7.4	(15,31)
	Luteolin	5280445	–6.1	–7.3	–7.3	–7.4	(15,32)
03	Saquinavir	441243	–7.3	–7.6	–8.1	–8.6	(33,34)
	Lopinavir	92727	–6.0	–6.9	–8.0	–7.6	(35−37)
	Ritonavir	392622	–6.8	–6.8	–7.2	–7.5	(35,37)
	Piperaquine	122262	–5.7	–7.5	–8.2	–7.4	(38)
	Dolutegravir	54726191	–6.6	–7.7	–7.8	–8.1	(39)
04	Fulvic Acid	5359407	–5.4	–7.0	–6.5	–7.0	(40,41)
	Humic Acid	90472028	–4.6	–5.0	–4.6	–5.9	(41)
	Piperine	638024	–5.5	–6.4	–6.8	–6.5	(42,43)
	Rose Oxide	27866	–4.1	–4.2	–4.6	–5.8	(44)
	Terpineol	17100	–5.3	–4.5	–5.3	–5.8	(45)
	Ferulate	445858	–4.5	–5.4	–5.6	–5.7	(46)
	Epicatechin	72276	–5.4	–7.0	–6.8	–6.6	(47−49)
	Capsaicin	1548943	–5.9	–5.7	–5.9	–5.4	(50,51)
	Arachidonic Acid	444899	–5.4	–5.0	–5.2	–5.1	(49)
	Kavain	5281565	–5.1	–5.7	–6.0	–6.2	(52)

Table shows that Receptors 1, 2, and 3 were obtained from the Protein Data Bank (PDB) (https://www.rcsb.org/).[16] Receptor 4 was built by homology modeling and obtained from ref (15).

Table 1

Information on SARS-CoV-2 Receptors Used

receptor	PDB code/reference	description	active site region/interface region
1	6W4B:A	The crystal structure of Nsp9 RNA binding protein of SARS CoV-2. Nsp9 replicase protein. Deposit Date: 2020–03–10.	Lys85
2	6Y84:A	SARS-CoV-2 main protease with unliganded active site (2019-nCoV, coronavirus disease 2019, COVID-19). Deposit Date: 2020–03–03.	Asn142
3	6VWW:A	Crystal structure of NSP15 endoribonuclease from SARS CoV-2. NSP15 endoribonuclease. Deposit Date: 2020–02–20.	His250
4	(15)	Computational model of the spike protein (S-protein) of SARS-CoV-2 interacting with the human ACE2 receptor. Publication date: 2020–02–27.	(15)

Molecular Docking

Molecular docking calculations were performed using AutoDock Vina software (https://vina.scripps.edu/).[17] The ligands and proteins were prepared for the calculations with AutoDock Tools (ADT) 1.5.6.[18] Hydrogens were added to both receptors and ligands individually. Gasteiger charges were then calculated by ADT, and nonpolar hydrogens were merged. The size of the grid box was set to 22.5 Å for each axis. The grid boxes were centered on the coordinates of the atoms of residues located in the region of the active site and interface region, as shown in Table . The number of modes was set to 50, and the exhaustiveness was set to 24. A total of 96 molecular docking calculations were performed. Binding energy (ΔGbind) values better or equal to −7 kcal/mol were established as a criterion for the efficiency of the interaction.

Results and Discussion

Table shows the results of the molecular docking calculations and references related to the ligands. The ligands in Groups 1 and 3 were selected for a drug repurposing study for COVID-19, the ligands in Group 2 were selected to analyze the interaction in the receptors of this study, and the ligands in Group 4 are chemical structures of natural compounds.

Regarding the ligands of Group 1, Table shows that hydroxychloroquine and azithromycin did not interact with the receptors, which may indicate that these drugs act in different steps in the viral cycle other than the ones that involve the proteins in this study. The literature cites three probable mechanisms of action for hydroxychloroquine: conformational modification of angiotensin-converting enzyme type 2 (ACE 2), alteration of the pH of the endosome, and reduction and inhibition in the release of pro-inflammatory cytokines (TNF-α and IL-6). Mefloquine showed interaction at Receptors 2 (−7.6 kcal/mol) and 3 (−7.3 kcal/mol). Nelfinavir and PL-69 showed interaction in three of the four studied receptors. Metaquine (N,N′-Bis(7-chloro-4-quinolyl)-m-phenylenediamine), mentioned in the literature as an antimalarial substance,[26] presented relevant results of interaction in all receptors, mainly in Receptor 3, where the binding energy was equal to −8.6 kcal/mol (see Figure ).

Figure 3

Global structure of the conformations resulted from docking. (A) Surface representation. (B) Representation of the region of the active site enlarged in transparent ribbons. His250 red, Ser294 green, Metaquine CPK, by sticks. (C) LigPlot+ diagram of the Receptor 3_Metaquine interaction.

The ligands in Group 02 showed relevant results in receptor 4 according to the study the of Smith and Smith, 2020.[15] In this study, Pemirolast and Benserazide ligands did not show any relevant interaction with the other three receptors. Luteolin showed an interaction with a binding energy value of −7.3 kcal/mol with Receptors 2 and 3. The results obtained for the three ligands at Receptor 4 (−7.4 kcal/mol) are in accordance with the study of Smith and Smith, 2020.[15]

Among Group 3 ligands, Lopinavir and Ritonavir showed interaction at two of the four receptors, Piperaquine and Dolutegravir had better results and showed interactions with three receptors. Saquinavir showed interaction with all receptors, with highlights to binding energy value of −8.6 kcal/mol at Receptor 4 (see Figure ).

Figure 4

Global structure of the conformations results from docking. (A) Surface representation. (B) Representation in ribbons. Asn33 yellow, Glu37 pink, Phe390 orange, Lys417 cyan, Saquinavir CPK, by sticks. (C) LigPlot+ diagram of the Receptor 4_Saquinavir interaction.

Regarding the ligands in Group 4, these did not present relevant results. What stood out the most in this group was Fulvic Acid, with a result of binding energy equal to −7 kcal/mol with Receptors 2 and 4. Epicatechin showed the same result with Receptor 2.

The results show that Saquinavir (Group 1) and Metaquine (Group 3) showed interaction with all the receptors in this study, indicating that these two drugs can be repurposed for the treatment of COVID-19. Nelfinavir and PL-69 (Group 1), Luteolin (Group 2), and Piperaquine and Dolutegravir (Group 3) interacted with three receptors and cannot be excluded as multitarget drug candidates for COVID-19. In silico studies can be carried out by changing the geometry of these ligands in order to improve the binding energy in the studied receptors; besides, in vitro and in vivo studies must be carried out to make promising drug candidates.

Metaquine is a potent, soluble, and bioavailable antimalarial substance. Metaquin shows activity in vivo (oral ID50 of 25 μmol/kg) against Plasmodium berghei and in vitro (0.17 μM) against Plasmodium falciparum K1 multidrug-resistant. Metaquin shows strong affinity for the putative heme antimalarial receptor.[26] Fielding et al.[27] showed that the inclusion of trifluoromethyl (CF3) in the chemical structure of Metaquine had a great impact on the drug interaction with the heme antimalarial receptor.

Saquinavir is a specific inhibitor drug for HIV protease. Jayaswal et al.[34] carried out an in silico study with modified Saquinavir structures to test the interaction with the HIV protease active site, with the aim of finding structures that present better binding energy than Saquinavir. Khadim et al.[33] conducted an in silico study of the interaction of SARS-Cov-2 main protease (Mpro) with anti-HIV drugs. The results obtained are in agreement with our study indicating that Saquinavir presented interactions with Mpro.

Figures –4 show information about the complexes with the best binding energy for each receptor. Regarding the LigPlot+ diagrams, Figures C, 2C, 3C, and 4C, the hydrogen bonds are represented by green dotted lines and the hydrophobic contacts are represented by red semicircles. The value that appears on the hydrogen bonds corresponds to the distance, in angstroms, between the amino acid residue of the protein and the ligand. Figures S1, S2, and S3 show interaction of Saquinavir at Receptors 1, 2, and 3, respectively, and Figure S4 shows the interaction of Metaquine at Receptor 4.

Figure 1

Global structure of the conformations resulted from docking. (A) Surface representation. (B) Representation in ribbons. Asn28 orange, Thr78 green, Lys85 red, Metaquine CPK, in sticks. (C) LigPlot+ diagram of the Receptor 1_Metaquine interaction [ref (53)]. Figure was generated using UCSF Chimera (https://www.cgl.ucsf.edu/chimera/) [ref (54)].

Figure 2

Global structure of the conformations resulted from docking. (A) Surface representation. (B) Representation of the region of the active site enlarged in transparent ribbons. Asn142 red, Metaquine CPK, by sticks. (C) LigPlot+ diagram of the Receptor 2_Metaquine interaction.

Figure C shows that Metaquine performed two interactions by hydrogen bonds with Receptor 1 (residues Asn28 and Thr78) and 11 interactions by hydrophobic contacts (residues Phe76, Val111, Leu46, Leu30, Asp27, Ala29, Asp48, Thr80, Ser47, Lys87, and Leu107).

Figure A shows the conformation of Metaquine in a cavity of Receptor 2 near the Asn142 residue (in red) in the region of the active site (see Table ). In Figure C, it is shown that Metaquine interacted by hydrophobic contacts with 15 residues (Asn142–active site, Glu166, Phe140, Leu141, Leu27, His163, Gly143, Cys145, Thr25, Thr45, His41, Met49, Ser46, Gln189, and Cys44).

Figure A shows the conformation of the Metaquine in a cavity of Receptor 3 near His250 (in red) in the region of the active site.

In Receptor 3, Metaquine performed an interaction by hydrogen bond (Ser294) and nine interactions by hydrophobic contacts (Val292, Lys290, His250–active site, Thr341, Trp333, His235, Tyr343, Cys293, and Leu346). Figure shows a part of the model created for human ACE2 receptor in green and part of the model created for spike protein (S-protein) in blue.

These parts of the two models were selected for molecular docking studies (see ref (15)). As it can be seen in Figure A, Saquinavir is positioned in the expected region, between the ACE2 receptor and the S-protein. Figure C shows that Saquinavir performed four interactions by hydrogen bond, three of them with the ACE2 receptor (Asn33, Glu37, and Phe390) and one with the S-protein (Lys417). In addition, it performed 18 interactions by hydrophobic contact with both models.

Conclusions

The results confirmed the interaction of Metaquine (N,N′-Bis(7-chloro-4-quinolyl)-m-phenylenediamine) and Saquinavir in all receptors studied, indicating them as candidates for multitarget drugs that can be repurposed to the treatment of COVID-19. In vitro, in vivo, and clinical tests should be performed to confirm the effectiveness of these drugs in the treatment of COVID-19.