A Bioluminescent 3CLPro Activity Assay to Monitor SARS-CoV-2 Replication and Identify Inhibitors

Abstract

Our therapeutic arsenal against viruses is very limited and the current pandemic of SARS-CoV-2 highlights the critical need for effective antivirals against emerging coronaviruses. Cellular assays allowing a precise quantification of viral replication in high-throughput experimental settings are essential to the screening of chemical libraries and the selection of best antiviral chemical structures. To develop a reporting system for SARS-CoV-2 infection, we generated cell lines expressing a firefly luciferase maintained in an inactive form by a consensus cleavage site for the viral protease 3CL^Pro of coronaviruses, so that the luminescent biosensor is turned on upon 3CL^Pro expression or SARS-CoV-2 infection. This cellular assay was used to screen a metabolism-oriented library of 492 compounds to identify metabolic vulnerabilities of coronaviruses for developing innovative therapeutic strategies. In agreement with recent reports, inhibitors of pyrimidine biosynthesis were found to prevent SARS-CoV-2 replication. Among the top hits, we also identified the NADPH oxidase (NOX) inhibitor Setanaxib. The anti-SARS-CoV-2 activity of Setanaxib was further confirmed using ACE2-expressing human pulmonary cells Beas2B as well as human primary nasal epithelial cells. Altogether, these results validate our cell-based functional assay and the interest of screening libraries of different origins to identify inhibitors of SARS-CoV-2 for drug repurposing or development.

Keywords: BAY2402234; DHODH; IPPA-17-A04; NADPH oxidase; SARS-CoV-2; Setanaxib; Vidofludimus; antiviral; chemical screening.

Figure 1

Development of a luminescent biosensor to measure the protease activity of SARS-CoV-2 3CL^Pro. (A) Schematic representation of SARS-CoV-2 genome with corresponding opened reading frames. PL^Pro and 3CL^Pro cleavage sites within the non-structural polyprotein are indicated by blue and red arrows, respectively. (B) Schematic representation of the luminescent biosensor designed for detecting 3CL^Pro activity. Firefly luciferase N and C-terminal fragments 4-233 (D1) and 234-544 (D2) were circularly permuted, and linked together by a peptide containing the 3CL^Pro cleavage site. The reporter protein is stabilized by circularization thanks to the IntN and IntC domains, while the PEST sequence promotes the degradation of non-ligated proteins. Upon proteolytic cleavage, constrained domains are relaxed and fold into a functional luminescent enzyme. (C) List of 3CL^Pro cleavage sites that are present in SARS-CoV-2 polyprotein ORF1ab. The nsp9-10 junction that was used in the luminescent biosensor is colored in red. (D) Sequence of the nsP9-10 junction across different alpha and betacoronaviruses. MHV is for Murine Hepatitis Virus or murine coronavirus, PEDV is Porcine Epidemic Diarrhea Virus, TGEV is for Transmissible Gastroenteritis Coronavirus and FCoV is for feline coronavirus.

Figure 2

Activation of the luminescent biosensor by SARS-CoV-2 3CL^Pro expression in different cell lines. (A–D) Caco-2, E6 Vero, Huh7 and Huh7.5 cells were stably transfected with a lentiviral vector to constitutively express the 3CL^Pro luminescent biosensor. Stable cell lines were transfected with an expression plasmid that was either empty (None) or encoding SARS-CoV-2 3CL^Pro. After 24 h of culture, luciferase activity in culture wells was determined. Data correspond to means ± SEM from 3 independent experiments in triplicate and p values were determined by Student’s t-test.

Figure 3

Activation of the luminescent biosensor by SARS-CoV-2 infection. (A) To compare their susceptibility to infection, Huh7-Lunet and Huh7.5 cells were infected with a recombinant strain of SARS-CoV-2 expressing mNeonGreen as a reporter (MOI = 0.5). After 20 h of culture, infection was determined by fluorescence microscopy and representative images are presented. Scale bar is 300 μm. (B) In 96-well plates, Huh7.5 cells stably transfected with the 3CL^Pro biosensor were infected with increasing amounts of SARS-CoV-2 (strain 2019-nCoV/USA_WA1/2020) ranging from 156 to 20,000 PFUs/well (i.e., 0.004 to 0.5 MOIs). After 24 h of culture, luciferase activity was determined. Data correspond to means ± SEM from 3 independent experiments in triplicate (n = 1) or in duplicate (n = 2). Linear regression was determined for PFUs/well ranging from 625 to 20,000 (i.e., 0.0156 to 0.5 MOIs) where some luciferase signal was detected. (C) Same experiment as in (B) but cells were infected with measles virus. Data correspond to means ± SEM from six experiments.

Figure 4

Identification of SARS-CoV-2 inhibitors by screening a chemical library of metabolic modulators. (A) Filtering pipeline for the selection of hit compounds. A total of 492 compounds were tested at 50 μM on Huh7.5 expressing the luminescent biosensor for 3CL^Pro. 20,000 cells per well were treated with the drugs for 24 h and then infected with SARS-CoV-2 (strain 2019-nCoV/USA_WA1/2020; MOI = 0.5; “screen w/ virus”) or were left uninfected (“screen w/o virus”). After 24 h, luciferase activity was determined and results were normalized across plates using DMSO-treated, uninfected control wells as reference. Activation of the luminescent biosensor upon SARS-CoV-2 infection was determined by calculating the normalized luminescence ratio (NLR = “luminescence w/ virus” over “luminescence w/o virus”), and results were expressed as a percentage of inhibitory effect compared to the DMSO-treated, infected control wells. Compounds showing no sign of cellular toxicity (see Section 2.4 for details) and inhibiting by 90% or more the activation of the luminescent biosensor in SARS-CoV-2 infected wells were selected. (B) Inhibitory effect of the 276 compounds showing no sign of cellular toxicity. Those with a negative inhibitory effect actually increased the activation of the luminescent biosensor upon SARS-CoV-2 infection. (C) List of the molecules selected from the screen. (D) Inhibitory effect of Setanaxib on the activation of the 3CL^Pro biosensor by SARS-CoV-2. Experiment was conducted as in (A) but with increasing concentrations of Setanaxib and results were expressed as percentage of activity compared to the infected, untreated control. Data correspond to means ± SEM from one experiment in triplicate.

Figure 5

Inhibition of SARS-CoV-2 replication by Setanaxib. (A,B) Huh7.5 cells were treated for 24 h with Setanaxib (25 μM) or left untreated, and then infected with recombinant SARS-CoV-2 expressing mNeonGreen (MOI = 0.1). After 24 h of culture, infection was determined by fluorescence microscopy. A representative image is presented in (A). Scale bar is 100 μm. A total of six randomly selected microscopy fields from three independent culture wells (one experiment in triplicate) were analyzed for mNeonGreen expression. Fluorescence quantification was performed with ImageJ, and corresponding data are presented in (B). (C) Same as in (A) but cells were infected with recombinant SARS-CoV-2 expressing mNeonGreen at 0.004, 0.02 and 0.1 MOIs. After 24 h of culture, supernatants were harvested and SARS-CoV-2 RNAs were quantified by RT-qPCR. (D) Beas-2B cells expressing ACE2 were infected with SARS-CoV-2 expressing mNeonGreen (MOI = 0.004) and cultured without or with Setanaxib at 25 μM. After 48 h, infection was determined by fluorescence microscopy for mNeonGreen expression. Quantitative data correspond to the fluorescence signal from six independent culture wells obtained with ImageJ (two independent experiments in triplicate). (E) Same experiment as in (D) but culture supernatants were collected to quantify viral RNA by RT-qPCR. Data correspond to means ± SEM. Statistical significance was determined by Student’s t-test using Prism (GraphPad Software, San Diego, CA, USA) for all panels except (C) where two-way ANOVA was used.

Figure 6

Setanaxib is inhibiting SARS-CoV-2 infection in human primary nasal epithelial cells. Primary nasal epithelial cells from healthy donors were infected with SARS-CoV-2 expressing mNeonGreen (MOI = 0.7) and cultured without or with Setanaxib at 12.5 or 25 μM. After 24 h, infection was determined by fluorescence microscopy for mNeonGreen expression. Scale bar is 300 μm. Fluorescence signal from culture wells was determined with ImageJ. Results are expressed as percentage of the untreated control well.

Figure 7

Setanaxib does not affect 3CL^Pro activity or viral fusion. (A) Huh7.5 cells expressing the 3CL^Pro luminescent biosensor were transfected with an expression plasmid that was either empty (None) or encoding SARS-CoV-2 3CL^Pro as described in Figure 2. After 5 h of incubation, Setanaxib was added at 25 μM. After 24 h of culture, luciferase activity in culture wells was determined and normalized to cell viability determined using the CellTiter-Glo reagent. Data correspond to means ± SEM from 2 independent experiments in triplicate and statistical significance was determined by Student’s t-test. (B) The potential impact of Setanaxib on membrane fusion induced the S protein of SARS-CoV-2 was determined in a cell-to-cell fusion assay. HEK-293T cells expressing both the S protein and the α peptide of β-galactosidase were co-incubated with HEK-293T cells expressing ACE2 and the Ω peptide of β-galactosidase. Cells were co-cultured for 5 h with increasing concentrations of Setanaxib to permit fusion, and β-galactosidase activity was determined using a bioluminescent substrate. Data correspond to means ± SEM from one experiment in triplicate and statistical significance was determined by Student’s t-test.