Differing pan-coronavirus antiviral potency of boceprevir and GC376 in vitro despite discordant molecular docking predictions

Given the structural similarities of the viral enzymes of different coronaviruses (CoVs), we investigated the potency of the anti-SARS-CoV-2 agents boceprevir and GC376 for counteracting seasonal coronavirus infections. In contrast to previous findings that both boceprevir and GC376 are potent inhibitors of the main protease (Mpro) of SARS-CoV-2, we found that GC376 is much more effective than boceprevir in inhibiting SARS-CoV-2 and three seasonal CoVs (NL63, 229E, and OC43) in cell culture models. However, these results are discordant with a molecular docking analysis that suggested comparable affinity of boceprevir and GC376 for the different Mpro enzymes of the four CoVs. Collectively, our results support future development of GC376 but not boceprevir (although it is an FDA-approved antiviral medication) as a pan-coronavirus antiviral agent. Furthermore, we caution against overinterpretation of in silico data when developing antiviral therapies. Supplementary Information The online version contains supplementary material available at 10.1007/s00705-022-05369-y.

Bart Haagmans (Department of Viroscience, Erasmus MC). Cell lines were analyzed by genotyping and confirmed to be mycoplasma negative.

Virus production and inoculation re-infection assay
LLCMK-2 cells harboring the infectious NL63 were seeded into multi-well plates, culturing at 33 °C, with 5% CO2 for 5-7 days, over 50% of cells have cytopathic effect (CPE), then harvested NL63 particles by repeated freezing and thawing three times, filtered with 0.45 μm filters. Huh7 cells harboring the infectious OC43 or 229E were seeded into multi-well plates, culturing at 33 °C, with 5% CO2 for 4-6 days. When over 50% of cells have cytopathic effect (CPE), OC43 or 229E particles were harvested by repeated freezing and thawing three times, then filtered with 0.45 μm filters. SARS-CoV-2 stocks were produced as previously described [1]. In short, Vero-E6 cells harboring the infectious SARS-CoV-2 were seeded into multi-well plates and incubating the cells at 37 °C, with 5% CO2 for 72 hours. The culture supernatant was cleared by centrifugation and stored in aliquots at −80°C. Cells were seeded into multi-well plates and culture medium was discarded when cell confluence was approximately 80%, followed by twice washing with 1×PBS. Harvested

TCID50 assay
Viruses in the cultured cells the supernatant were harvested through repeated freezing and thawing for three times. NL63 titer was quantified by using a 50% tissue culture infectious dose (TCID50) assay. Briefly, ten-fold dilutions of NL63 were inoculated onto LLCMK-2 cells, grown in a 96-well tissue culture plate at 2,000 cells/well. The plate was incubated at 33 °C for 7 days, and each well was examined under a light microscope for cytopathic effect (CPE). The TCID50 value was calculated by using the Reed-Muench method.

Quantification of NL63 genome copy numbers
An amplicon of NL63 (a fragment of N protein) were cloned into the pCR2.1-TOPO vector (Invitrogen, San Diego, CA) to generate a template for quantifying NL63 genome copy numbers. The plasmid was extracted by Quick Plasmid Miniprep Kit (Invitrogen, Lohne, Germany). A series of dilutions (from 10 -1 to 10 -8 ) were prepared and then were amplified and quantified by qRT-PCR to generate a standard curve. This standard curve was generated by plotting the log copy number versus the cycle threshold (CT) value (Fig. S3).

3D protein structure modeling of coronavirus Mpro
The experimentally solved crystal structure of SARS-CoV-2 Mpro (PDB Id 6M2Q), NL63 Mpro (PDB Id 5GWY), and 229E Mpro (PDB Id 2ZU2) were retrieved from RCSB PDB database (https://www.rcsb.org). The retrieved structures were prepared for docking in Discovery Studio Visualizer. The hetero atoms, water molecules and any co-crystalized ligand groups were removed followed by addition of polar hydrogen atoms. The crystal structure for OC43 Mpro was not available at PDB database, therefore, we perform the homology-based 3D structure modelling in the Modeller 10.1 tool [2]. Briefly, the protein sequence of OC43 Mpro was submitted for structure-based similarity search in NCBI-BLAST using RCSB PDB as the search database to obtain best template for modelling and HKU1 Mpro (PDB ID: 3D23) showed a percent identity of 82% with the sequence of OC43 Mpro as well as 99% query cover for structure modelling. A total of 50 models were prepared in Modeller amino acids in the most favored region while no residues in disallowed region and an ERRAT (quality factor) score of 97%, demonstrating a good model structure for further study. Lastly, the root mean square deviation was also computed between the template (HKU1 Mpro) and the OC43 modelled structure by superimposing in Chimera tool and we obtained a RMSD value of 0.398Å [3] (Fig. S7).

Protein-ligand docking
Boceprevir (CID 10324367) and GC-376 (CID 71481120) retrieved from PubChem database were prepared for molecular docking with Mpro structures of four coronaviruses. The protein-ligand docking was performed in AutoDock Vina and top most pose (1st docking pose) with highest binding energy (in kcal/mol) was opted for further analysis [4]. In addition to binding energy, it was very important to analyze the interaction of both drugs with residues of Mpro, mainly their non-covalent intermolecular interaction in the catalytic target site.
The interaction analysis between drugs and protein was performed in Discovery studio visualizer.

Statistics analysis
All numerical results were reported as Mean ± SEM. The statistical significance of differences between means was assessed with the Mann-Whitney test (GraphPad Prism 5; GraphPad Software Inc., La Jolla, CA).
The threshold for statistical significance was defined as P ≤ 0.05. between template (blue) and model structure (purple) was computed in chimera tool as 0.398Å only, (B) An ERRAT score of 97% provides an overall quality factor, and (C) The Ramachadaran plot shows 95.5% amino acid in most favored region while no residues in disallowed region stereo chemically signifies the quality of OC43 Mpro modelled structure.