key: cord-0911149-txrcs0d7
authors: Pitsillou, Eleni; Liang, Julia; Karagiannis, Chris; Ververis, Katherine; Darmawan, Kevion K.; Ng, Ken; Hung, Andrew; Karagiannis, Tom C.
title: Interaction of small molecules with the SARS-CoV-2 main protease in silico and in vitro validation of potential lead compounds using an enzyme-linked immunosorbent assay
date: 2020-10-23
journal: Comput Biol Chem
DOI: 10.1016/j.compbiolchem.2020.107408
sha: 5476b7248e0b63d130f8361aee03593bbc229d8e
doc_id: 911149
cord_uid: txrcs0d7

Caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the COVID-19 pandemic is ongoing, with no proven safe and effective vaccine to date. Further, effective therapeutic agents for COVID-19 are limited, and as a result, the identification of potential small molecule antiviral drugs is of particular importance. A critical antiviral target is the SARS-CoV-2 main protease (M(pro)), and our aim was to identify lead compounds with potential inhibitory effects. We performed an initial molecular docking screen of 300 small molecules, which included phenolic compounds and fatty acids from our OliveNet™ library (224), and an additional group of curated pharmacological and dietary compounds. The prototypical α-ketoamide 13b inhibitor was used as a control to guide selection of the top 30 compounds with respect to binding affinity to the M(pro) active site. Further studies and analyses including blind docking were performed to identify hypericin, cyanidin-3-O-glucoside and SRT2104 as potential leads. Molecular dynamics simulations demonstrated that hypericin (ΔG = -18.6 and -19.3 kcal/mol), cyanidin-3-O-glucoside (ΔG = -50.8 and -42.1 kcal/mol), and SRT2104 (ΔG = -8.7 and -20.6 kcal/mol), formed stable interactions with the M(pro) active site. An enzyme-linked immunosorbent assay indicated that, albeit, not as potent as the covalent positive control (GC376), our leads inhibited the M(pro) with activity in the micromolar range, and an order of effectiveness of hypericin and cyanidin-3-O-glucoside > SRT2104 > SRT1720. Overall, our findings, and those highlighted by others indicate that hypericin and cyanidin-3-O-glucoside are suitable candidates for progress to in vitro and in vivo antiviral studies.

Since being declared a pandemic in early March, COVID-19 has spread rapidly throughout the world and is currently ongoing (1) . There are currently no approved human coronavirus vaccines (2) . As a result, there is an urgent need to investigate, identify and repurpose small molecules with potential antiviral effects (3) . To date the RNA-dependent RNA polymerase J o u r n a l P r e -p r o o f inhibitor, remdesivir, is the only antiviral drug that has been approved; it has been granted emergency use authorisation for the compassionate management of severe COVID-19 (4) .

In addition to the spike glycoprotein, the coronavirus main protease (M pro ) is an important target for antiviral therapy (5) . The replicase gene of SARS-CoV-2 encodes the replicase polyproteins, pp1a and pp1ab (5) (6) (7) . M pro is the enzyme that is predominantly responsible for the proteolytic processing of these polyproteins into functional polypeptides (5) (6) (7) . Thus, M pro plays an important role in viral replication and infection (6) . There is a growing body of literature on lead compounds that could be used to target the M pro enzyme of SARS-CoV-2.

This includes peptidomimetics, such as -ketoamide inhibitors (6, 8, 9) . Aside from synthetic ligands, researchers are also investigating the antiviral properties of natural compounds and dietary polyphenols have been of particular interest (5, (10) (11) (12) (13) .

Extra-virgin olive oil (EVOO), which is the primary source of dietary fat within the Mediterranean diet, is rich in phenolic compounds (14) . The OliveNet TM library, which was previously created by our laboratory, is a curated database of 676 compounds from Olea europaea and 222 phenolic compounds are divided into 13 subclasses (15) . Crystal structures of the SARS-CoV-2 M pro enzyme have been made available on the RCSB Protein Data Bank (PDB) and can be used for in silico screening (6, 9) . Utilising a combination of targeted molecular docking and blind docking, we aimed to identify 'hit' compounds from a selection of 300 ligands that could potentially inhibit M pro . This included 211 phenolic compounds and 13 fatty acids from our OliveNet TM library, known protease inhibitors, several antibiotics for comparison and the -ketoamide inhibitor as a control (15) . Molecular dynamics (MD) simulations were subsequently performed to evaluate the top three candidates and ultimately determine the lead compounds.

The crystal structures of SARS-CoV-2 M pro was obtained from the RSCB Protein Data Bank (PDB ID: 6LU7, 6Y2G, 6M03) (9, (16) (17) (18) . 300 compounds were selected for screening against M pro .. This comprised of 211 phenolic compounds and 13 fatty acids sourced from the OliveNet TM Library (15) , and an additional 76 ligands based on known protease inhibitors and antibiotics, as well as compounds with antiviral, anti-inflammatory, anti-parasitic, antimalarial, antioxidant and anti-ageing properties (8, 11, . A full list of the ligands that were screened can be found in the supplementary information (Table S1 ). Ligand structures were obtained from the National Centre for Biotechnology Information (NCBI) PubChem database (44) , or drawn using Chem3D 19.0 (Perkin Elmer, Massachusetts, USA) if they weren't available.

Structure preparation and molecular docking was performed using the quantum-mechanicspolarised ligand docking (QPLD) protocol of the Schrödinger Suite (version 2018-1 and 2020-2) molecular modelling package (45-49) as previously described (50) . A 20 x 20 x 20 Å receptor grid was generated centred around active site residues Thr24, Thr25, His163, Pro168 and Gly143 (6) .

Compounds were also docked to the active site of M pro using AutoDock Vina (51), following the processing of protein and ligand structures using PyRx (52) to generate their corresponding pdbqt files. The protein structure was assumed to be rigid, and rotatable torsions of the ligands were activated. A receptor grid with dimensions of 25 x 25 x 25 Å was generated around the same active site residues. Docking was performed with an exhaustiveness of 128.

J o u r n a l P r e -p r o o f Docking calculations were performed on a Windows 10 workstation equipped with an Intel Core i7 (2.90 GHz) and 8.00 GB of RAM.

As the SARS-CoV-2 M pro is known to function as a homodimer (16) , this complex was assembled using the Proteins, Interfaces, Structures and Assemblies (PDBePISA) server (53) for blind docking to identify potential binding sites. Structures were processed in PyRx, and docking performed AutoDock Vina (51) using a receptor grid encompassing the entire protein surface at an exhaustiveness of 128. For selected top binding compounds, blind docking was also performed at an exhaustiveness of 2000 using cloud computing services provided by Galileo (Hypernet Labs).

Classical MD simulations were performed using GROMACS 2018.2 software (54, 55) with the CHARMM27 force field (56, 57) using docked ligands as starting structures as previously described (50) . Ligand topologies were generated using SwissParam (58) . For cyanidin-3-Oglucoside, Lennard-Jones parameters for the oxonium ion were assumed to be similar to those for the ether group (59) . Simulations were performed with a time-step of 2 fs in triplicate for 100 ns. 

To confirm inhibition of the SARS-CoV-2 M pro in vitro, an enzyme-linked immunosorbent assay (ELISA), was performed. The BPL 3CL protease (SARS-CoV-2) assay kit (BPS Bioscience, San Diego, CA, USA), was used, and the assay performed according to the manufacturer's instructions. The internal positive control was the broad-spectrum antiviral determinations) into account. The IC50 values for applicable test inhibitors (hypericin, cyanidin-3-O-glucoside, and SRT2104), were also calculated.

Identification of key compounds with relatively high affinity to the active site of the SARS-CoV-2 main protease J o u r n a l P r e -p r o o f

The -ketoamide ligand (13b) was previously identified by Zhang et al. to inhibit M pro with a half maximal inhibitory concentration (IC50) value of 0.67 ± 0.18 M (9). As a result, this compound was used as a control. Each protomer of the M pro enzyme consists of three domains and the active site is located between domain I and domain II (6) . When docked to the substratebinding site of the crystal structure, the -ketoamide ligand formed hydrogen bonds with the protein residues and this included Glu166, His164 and Gln189. The inhibitor was predominantly surrounded by hydrophobic and polar residues including Cys145, Asn142, Tyr54, Thr190 and Pro168 ( Figure 1 ). The binding affinities of this compound were -65.7 and -7.7 kcal/mol in Schrödinger and AutoDock Vina, respectively. Our lab has also recently verified the interaction of the ketoamide inhibitor with the active site of M pro from SARS-CoV-2 (50).

The control compound was positioned in a similar site as the co-crystallised ligand (N3) (6).

Previously, it has been found that the negatively charged residue Glu166 plays an important role in forming the S1 pocket of the binding site (6) . The hydrophobic residue Cys145 is also involved in the mechanisms of action of N3 and the -ketoamide inhibitor (6, 9) . The Cys145 and His41 residues in the SARS-CoV-2 main protease form a catalytic dyad (6) . The M pro enzyme is a cysteine protease and the inhibitors specifically interact with Cys145 covalently (6, 9) . Covalent docking tools have been made available however, the success of this screening approach depends on a number of factors (65, 66) . This includes the contribution of noncovalent interactions and the mechanism of covalent bonding (65, 66) . In the current study, conventional docking methods were used and there is evidence to suggest that noncovalent docking is successful in elucidating the interactions that are occurring within protein-ligand complexes at the molecular level (67-69).

Drug repositioning has become one of the most important strategies for combating and virtual screening approaches continue to play a major role in this (70) . With this in mind, 300 compounds were docked to the catalytic core of the SARS-CoV-2 M pro enzyme. Binding to the active site of the protomer using the QPLD protocol in Schrödinger yielded Glide energies ranging from -16.3 to -82.0 kcal/mol. All 300 compounds were predicted to bind using AutoDock Vina and the binding affinities ranged from -3.7 to -10.7 kcal/mol. When examining the effect of molecular weight of the small molecules and when comparing the binding affinities from both programs, the correlation coefficients were found to be approximately 0. In regards to the 211 phenolic compounds from the OliveNet TM database, it was evident that the flavonoid, glucoside and secoiridoid subclasses were binding strongly to the active site (Table S1 ). Conversely, the simple phenols, hydroxyphenylacetic acids, hydroxybenzoic acids and methoxyphenols had weaker binding affinities (Table S1 ). The biological activities of flavonoids have been extensively investigated over the years and there are studies that have examined the inhibitory activity of certain flavonoid compounds against the M pro enzyme of SARS-CoV. This includes luteolin, tetra-O-galloyl--D-glucose and baicalin to name a few (13, 72, 73) . Since being declared a pandemic, several papers that have assessed the ability of natural compounds to target SARS-CoV-2 proteins have been made available (74, 75) . In a J o u r n a l P r e -p r o o f recent study conducted by Ul Qamar et al. in silico techniques were used to detect lead molecules from a medical plant library and they highlighted how their study may contribute to the development of natural antiviral agents in the future (76) . Although there are hurdles that are yet to be overcome, Thomford et al. emphasise that advancements in predictive computational methods have made it possible for the properties of natural products and their derivatives to be explored and for novel therapeutic moieties to be discovered (77) .

Saquinavir and ritonavir were the protease inhibitors that were binding more strongly than the -ketoamide ligand, while nelfinavir had a similar Glide energy as the control compound. This is in accordance with a paper published by Pant et al., as their molecular docking analysis revealed that these three antivirals scored well and interacted with important residues (78) . In addition to the phenolic compounds from OliveNet TM and the protease inhibitors, some of the other top binding ligands included (-)-epicatechin gallate, remdesivir, D,L-sulforaphane glutathione, SRT2104, SRT1720, hypericin, curcumin, demethoxycurcumin, baricitinib and baicalin.

Based on this initial screen, 30 compounds were selected for further examination. To select the 30 compounds from the initial screen of the 300 compounds, a number of criteria were applied.

Firstly, we selected compounds on the basis of binding affinity; compounds that were significantly below the positive controls, in this case, α-ketoamide 13b, were the first group to be eliminated. We then took into account compound availability. Given that we are trying to identify potential lead compounds for the current ongoing pandemic, it is important that the compounds are at least commercially available, thus allowing potential further investigation in vitro and in vivo. Many of our compounds would require synthesis and a complete evaluation of bioactivity before they can used in further experiments. Therefore, the more obscure lesswell characterized compounds were the next group to be eliminated. It should be noted that J o u r n a l P r e -p r o o f the screening data obtained for these less well-studied compounds would still be useful, for potential structure activity-related studies, and for potential synthesis and evaluation at a more appropriate time. Compounds with relatively high binding affinity for the active site, that have been studied in animal models, and preferably in humans were the first to be selected for further evaluation. A description of these compounds and their binding affinities can be found in Table   1 . The structures of the ligands not found in the OliveNet TM database are provided in the supplementary information (Table S2 ).

In order to narrow down the list and identify lead compounds, blind docking was conducted on 

To assess the stability of the ligands in complex with the main protease dimer, classical MD simulations were performed. Each system comprised of two ligands bound to the active sites on each protomer, as shown in Movies S1-4. Root mean square deviation (RMSD) analysis in In contrast to hypericin and SRT2104, residue energy contributions for cyanidin-3-O-glucoside demonstrates large fluctuations across the entire protein, regardless of the protomer to which the ligand is bound. Energy peaks were nevertheless present at active site residues Ser1, Asp187, and Arg188. Peaks for Glu166 across both protomers were conversely favourable contributions. While more rigorous free energy methods may be required for further investigation, the dominating favourable energy contributions yield a strong ΔG of cyanidin-3-O-glucoside to the active site of M pro , which is observed to remain stable throughout the trajectory (Movie S4).

Cyanidin-3-O-glucoside belongs to the flavonoid subclass and is classified as an anthocyanin (80) . In general, anthocyanins are water-soluble pigments and they are present in a variety of plants and fruits (80, 81) . Their antioxidant, anti-inflammatory and neuroprotective properties have gained the attention of researchers and different methodologies are being trialled to assess how the bioavailability of these compounds could be increased (12, 82, 83) . Pour et al. have also published a comprehensive review on the antiviral properties of anthocyanins and emphasised the need for novel drugs to be rapidly developed (12) . We also identified the sirtuin 1 (SIRT1) activator SRT2104 as a potential hit ligand (84) .

Furthermore, hypericin is a compound that is naturally found in the perennial plant Hypericum perforatum (St. John's wort) and its mechanisms of action require further elucidation (85) (86) (87) (88) .

Hypericin is considered to be a potent photosensitising agent and its potential use in cancer therapy has been investigated (88) . The antidepressant effects of St. John's wort have also been reported in the literature (86) . Likewise, the protective properties of hypericin against J o u r n a l P r e -p r o o f enveloped and non-enveloped viruses have been explored and previous studies have focused on hepatitis C, infectious bronchitis virus and human immunodeficiency virus 1 (HIV-1) (89) (90) (91) (92) (93) (94) . It is important to note that drug-drug interactions have been described for hypericin and medications such as HIV protease inhibitors (95, 96) . Therefore, it is imperative that the pharmacokinetics of these compounds are taken into consideration. Hypericin and cyanidin-3-O-glucoside were consequently identified as the lead dietary compounds in this study and this is also consistent with a recently published paper by Islam et al (97) .

To confirm potential inhibition of the SARS-CoV-2 in vitro, we performed an ELISA using a commercially available 3CL protease (SARS-CoV-2) assay kit. The IC50 for the GC376 positive antiviral control used in this ELISA has been calculated to be 0.46 µM (as provided the kit supplier; BPS Bioscience). Our findings indicate that hypericin results in a concentration-dependent inhibition of M pro activity, with an IC50 value calculated to be 63.6 ± 5.7 µM ( Figure 6 and Table 2 Table 2 ). Notably, IC50 values for SRT1720, resveratrol and L-sulforaphane could not be determined (up to 128 µM, Table 2 ), consistent with the lower binding energies observed for these compounds in the in silico work (Table S1) . Similarly, the percentage protease inhibition at 50 µM GC376 was determined to be 97.9 ± 1.8% in our experiments (n = 9 determinations, Figure 6 ). As shown in Table 2 

COVID-19): Situation Report -51. Geneva: World Health Organization

SARS-CoV-2 vaccines: status report

Clinical trials on drug repositioning for COVID-19 treatment

Hints of hope with remdesivir

An overview of severe acute respiratory syndrome-coronavirus (SARS-CoV) 3CL protease inhibitors: peptidomimetics and small molecule chemotherapy

A novel coronavirus from patients with pneumonia in China

Structures of two coronavirus main proteases: implications for substrate binding and antiviral drug design

Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2

Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved α-ketoamide inhibitors

Antiviral activity of plants and their isolated bioactive compounds: An update

Flavonoids: promising natural compounds against viral infections

The signaling Pathways, and therapeutic targets of antiviral agents: focusing on the antiviral approaches and clinical perspectives of anthocyanins in the management of viral diseases

Small molecules blocking the entry of severe acute respiratory syndrome coronavirus into host cells

The Mediterranean diet: from an environment-driven food culture to an emerging medical prescription

OliveNet™: a comprehensive library of compounds from Olea europaea

Structure of Mpro from COVID-19 virus and discovery of its inhibitors

The Protein Data Bank

The crystal structure of COVID-19 main protease in apo form

Proteases and protease inhibitors in infectious diseases

Coronavirus susceptibility to the antiviral remdesivir (GS-5734) is mediated by the viral polymerase and the proofreading exoribonuclease. mBio

Indomethacin has a potent antiviral activity against SARS coronavirus

Potent inhibition of HIV-1 replication in resting CD4 T cells by resveratrol and pterostilbene

Antiviral activity against infectious bronchitis virus and bioactive components of Hypericum perforatum L

Pharmacokinetics of oseltamivir: an oral antiviral for the treatment and prophylaxis of influenza in diverse populations

Should chloroquine and hydroxychloroquine be used to treat COVID-19? A rapid review

Possible immunosuppressive effects of drug exposure and environmental and nutritional effects on infection and vaccination

Sirtuin 1 activator SRT1720 suppresses inflammation in an ovalbumin-induced mouse model of asthma

Potential inhibitor of COVID-19 main protease (Mpro) from several medicinal plant compounds by molecular docking study2020

Natural bisbenzylisoquinoline alkaloids-tetrandrine, fangchinoline, and cepharanthine, inhibit human coronavirus OC43 infection of MRC-5 human lung cells

high-throughput screening of natural compounds of MERS-CoV entry inhibitors using a pseudovirus expressing MERS-CoV spike protein

A randomized, placebo-controlled study of SRT2104, a SIRT1 activator, in patients with moderate to severe psoriasis

Investigation of potential anti-pneumococcal effects of l-sulforaphane and metabolites: Insights from synchrotron-FTIR microspectroscopy and molecular docking studies

Disulfiram can inhibit MERS and SARS coronavirus papain-like proteases via different modes

Effective inhibition of MERS-CoV infection by resveratrol

Synergistic effects of combination treatment using EGCG and suramin against the chikungunya virus

A review on antibacterial, antiviral, and antifungal activity of curcumin

Baicalin, a metabolite of baicalein with antiviral activity against dengue virus

The signaling pathways, and therapeutic targets of antiviral agents: focusing on the antiviral approaches and clinical perspectives of anthocyanins in the management of viral diseases

Baricitinib as potential treatment for 2019-nCoV acute respiratory disease. The Lancet

Cepharanthine: a review of the antiviral potential of a Japanese-approved alopecia drug in COVID-19. Pharmacological Reports. 2020. 41. Vankadari N. Arbidol: A potential antiviral drug for the treatment of SARS-CoV-2 by blocking trimerization of the spike glycoprotein

Potential therapeutic targets and promising drugs for combating SARS-CoV-2

Network-based drug repurposing for novel coronavirus 2019-nCoV/SARS-CoV-2

PubChem 2019 update: improved access to chemical data

Protein and ligand preparation: parameters, protocols, and influence on virtual screening enrichments

An interactive web-based dashboard to track COVID-19 in real time

Intensive care management of coronavirus disease 2019 (COVID-19): challenges and recommendations

Importance of accurate charges in molecular docking: Quantum mechanical/molecular mechanical (QM/MM) approach

Interaction of the prototypical α-ketoamide inhibitor with the SARS-CoV-2 main protease active site in silico: Molecular dynamic simulations highlight the stability of the ligand-protein complex

AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading

Small-molecule library screening by docking with PyRx

Protein interfaces, surfaces and assemblies service PISA at European Bioinformatics Institute

A message-passing parallel molecular dynamics implementation

High performance molecular simulations through multi-level parallelism from laptops to supercomputers

Implementation of the CHARMM Force Field in GROMACS: Analysis of Protein Stability Effects from Correction Maps, Virtual Interaction Sites, and Water Models

A force field for drug-like molecules compatible with the CHARMM all-atom additive biological force fields

SwissParam: a fast force field generation tool for small organic molecules

Structure and dynamics of hydronium in the ion channel gramicidin A

Electrostatics of nanosystems: Application to microtubules and the ribosome

g_mmpbsa-A GROMACS Tool for High-Throughput MM-PBSA Calculations

Assessing the Performance of the MM/PBSA and MM/GBSA Methods. 1. The Accuracy of Binding Free Energy Calculations Based on Molecular Dynamics Simulations

VMD: visual molecular dynamics

Spartan HPC-Cloud Hybrid: Delivering Performance and Flexibility2017

Merits and pitfalls of conventional and covalent docking in identifying new hydroxyl aryl aldehyde like compounds as human IRE1 inhibitors

Comparative evaluation of covalent docking tools

Molecular docking: challenges, advances and its use in drug discovery perspective

Molecular docking: a powerful approach for structure-based drug discovery

Comprehensive evaluation of ten docking programs on a diverse set of protein-ligand complexes: the prediction accuracy of sampling power and scoring power

Current status of epidemiology, diagnosis, therapeutics, and vaccines for novel Coronavirus Disease 2019 (COVID-19)

A detailed comparison of current docking and scoring methods on systems of pharmaceutical relevance

In vitro susceptibility of 10 clinical isolates of SARS coronavirus to selected antiviral compounds

Severe acute respiratory syndrome

Natural-like products as potential SARS-CoV-2 M(pro) inhibitors: in-silico drug discovery

In-Silico approach for identification of effective and stable inhibitors for COVID-19 main protease (M(pro)) from flavonoid based phytochemical constituents of Calendula officinalis

Structural basis of SARS-CoV

3CL(pro) and anti-COVID-19 drug discovery from medicinal plants

Natural products for drug discovery in the 21st century: innovations for novel drug discovery

Peptide-like and smallmolecule inhibitors against Covid-19

Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved alpha-ketoamide inhibitors

Anthocyanins and metabolites from purple rice inhibit IL-1β-induced matrix metalloproteinases expression in human articular chondrocytes through the NF-κB and ERK/MAPK pathway

Pomegranate fruit and juice (cv. Mollar), rich in ellagitannins and anthocyanins, also provide a significant content of a wide range of proanthocyanidins

Anthocyanins encapsulated by PLGA@PEG nanoparticles potentially improved its free radical scavenging capabilities via p38/JNK pathway against Aβ(1-42)-induced oxidative stress

The case for anthocyanin consumption to promote human health: a review

Pharmacokinetics and tolerability of SRT2104, a first-in-class small molecule activator of SIRT1, after single and repeated oral administration in man

St John's wort (Hypericum perforatum L.): a review of its chemistry, pharmacology and clinical properties

The mechanisms of action of St. John's wort: an update

Über das Hypericin, den photodynamisch wirksamen Farbstoff aus Hypericum perforatum

Hypericin in the light and in the dark: two sides of the same coin

The importance of light in the anti-HIV effect of hypericin

Hypericin inhibits hepatitis C virus replication via deacetylation and down-regulation of heme oxygenase-1

Protective effects of hypericin against infectious bronchitis virus induced apoptosis and reactive oxygen species in chicken embryo kidney cells

John's wort plant, in patients with chronic hepatitis C virus infection

Antiviral activity against infectious bronchitis virus and bioactive components of Hypericum perforatum L

Virucidal activity of hypericin against enveloped and non-enveloped DNA and RNA viruses

In vitro interaction of the HIV protease inhibitor ritonavir with herbal constituents: changes in P-gp and CYP3A4 activity

Phytotherapeutics: the emerging role of intestinal and hepatocellular transporters in dug interactions with botanical supplements

A molecular modeling approach to identify effective antiviral phytochemicals against the main protease of SARS-CoV-2

Hypericin (closed circles) and cyanidin-3-O-glucoside (open circles) were assayed in triplicate and the average ± SEM values are depicted. Concentration-dependent inhibition of the SARS-CoV-2 M pro by hypericin and cyanidin-3-O-glucoside. To confirm in vitro inhibition a 3CL protease ELISA assay was performed and fluorescence intensities at an emission of wavelength of 460 nm for concentrations of hypericin and cyanidin-3-O-glucoside up to 128 µM (doubling dilutions), were measured. The average background (n = 9 determinations), total enzymatic activity (n = 9 determinations), and inhibition by the covalent internal positive control

J o u r n a l P r e -p r o o f

J o u r n a l P r e -p r o o f