A simple reverse genetics method to generate recombinant coronaviruses

Abstract

Engineering recombinant viruses is a pre-eminent tool for deciphering the biology of emerging viral pathogens such as the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). However, the large size of coronavirus genomes renders the current reverse genetics methods challenging. Here, we describe a simple method based on "infectious subgenomic amplicons" (ISA) technology to generate recombinant infectious coronaviruses with no need for reconstruction of the complete genomic cDNA and apply this method to SARS-CoV-2 and also to the feline enteric coronavirus. In both cases we rescue wild-type viruses with biological characteristics similar to original strains. Specific mutations and fluorescent red reporter genes can be readily incorporated into the SARS-CoV-2 genome enabling the generation of a genomic variants and fluorescent reporter strains for in vivo experiments, serological diagnosis, and antiviral assays. The swiftness and simplicity of the ISA method has the potential to facilitate the advance of coronavirus reverse genetics studies, to explore the molecular biological properties of the SARS-CoV-2 variants, and to accelerate the development of effective therapeutic reagents.

Keywords: SARS-CoV-2; antivirals; in vivo experiment; reverse genetics; serology.

Figure 1. The ISA method to rescue SARS‐CoV‐2

SARS‐CoV‐2 complete genome sequence was used to design eight overlapping subgenomic viral fragments covering the complete genome. Positions on the genome (in nucleotide) are indicated in bold red. This figure was created with BioRender.com.

Figure EV1. The spike, nucleocapsid, membrane, and envelope proteins of clinical and recombinant SARS‐CoV‐2 after two passages in VeroE6 were amplified by RT‐PCR from clarified supernatant medium

Figure 2. Virus replication kinetics of clinical and ISA strains

1. A, B

   An moi of 0.001 was used to infect VeroE6 with rescued or clinical SARS‐CoV‐2 (A, B).

2. C, D

   An moi of 0.01 was used to infect FeA cells with rescued or clinical FeCoV (C, D).

Data information: Data are represented as mean ± SD (indicated by the error bars). Each experiment was performed in technical triplicates (N = 3). Exploratory analyses were performed using a two‐way ANOVA for multiple comparisons with Sidak’s multiple comparisons test. Statistical comparisons were performed between SARS‐CoV‐2 clinical European vs ISA European strains, ISA European vs ISA D614 strains, ISA D614 vs mCherry D614 strains, and between FeCoV clinical vs ISA strains. Only P‐values ≤ 0.05 were indicated by a * symbol. In other cases, P‐values > 0.05 were considered as not significant and were not displayed on the graph. ***, **, and * symbols indicate that the average value for the ISA D614 strain is significantly different from that of the ISA European strain with P‐values < 0. 0001, < 001, and ≤ 0.05, respectively.

Figure EV2. The ISA method to rescue FeCoV

FeCoV complete genome sequence was used to design eight overlapping subgenomic viral fragments covering the full genome. Positions on the genome (in nucleotide) are indicated in bold red. This figure was created with BioRender.com.

Figure 3. Fluorescence microscopy analysis and correlation between titers of neutralizing antibodies (nAb) using ISA D614 and mCherry D614 SARS‐CoV‐2 strains

1. Vero E6 cells were infected with an moi of 0.05 with the fluorescent mCherry D614 strains, wild‐type ISA D614 or mock infected. Pictures were taken at 48 h pi (20×). Scale bar, 100 µm.

2. The mCherry reporter gene in the ISA mCherry D614 on VeroE6 cells supernatant medium at passages 2 to 5 (p2, p3, p4, and p5) was RT‐PCR amplified and analyzed using gel electrophoresis.

3. Twenty‐four human sera were then two‐fold diluted and incubated with the ISA D614 and mCherry D614 strains and nAb titers were recorded at 5 days dpi. nAb titers were defined as the highest dilution that inhibited the production of distinct cpe with the ISA D614 SARS‐CoV‐2 or fluorescence with the fluorescent mCherry D614 SARS‐CoV‐2. Each black dot represents results from a given number of sera. Statistical analyses were performed using univariate linear regression. The error band (in grey) represents the 95% confidence interval of the regression line. The Pearson correlation coefficient (R ²) and P‐value analyses are shown.

4. Representative neutralizing curves of the nAb fluorescence‐based assay. The four‐parameter dose–response curve was fitted using the nonlinear regression method and nAbs were calculated in the software Prism 7.0. For negative serum samples, an arbitrary value of 10 was assigned (detection threshold for both methods).

Figure 4. Remdesivir antiviral activity on SARS‐CoV‐2 in VeroE6 cells

1. Dose–response curve for the ISA D614 and for the mCherry D614 strains obtained by fluorescence or viral RNA measurement in VeroE6 cells from one representative experiment. Data are represented as mean ± SD (indicated by the error bars). Each experiment was performed in technical triplicates (N = 3).

2. Table of EC₅₀ values obtained for the two different strains from two technical replicates and their respective mean ± SD.

3. Fluorescence of the SARS‐CoV‐2 mCherry in VeroE6 cells with different Remdesivir concentration. Scale bar, 200 µm.

Data information: EC₅₀: 50% inhibition, Remdesivir concentrations are presented in log scale for logarithmic interpolation. Dose–response curves were generated using GraphPad Prism software version 7.0 (https://graphpad‐prism.software.informer.com/7.0/) with a four‐parameter linear regression.

Figure 5. Body weight changes and viral replication in tissues after infection by SARS‐CoV‐2 in Syrian gold hamsters

Groups of four hamsters were intranasally infected with 10³ TCID₅₀ of clinical European, ISA European or ISA D614 strain. After 3 dpi, viral RNA loads and infectious viral loads were assessed in lung and plasma.

1. Clinical course of the disease. Normalized weight at day n was calculated as follows: % of initial weight of the animal at day n.

2. Lung infectious titers (measured using a TCID₅₀ assay) expressed in TCID_50/g of lung.

3. Lung viral RNA yields (measured using an RT‐qPCR assay) expressed in virus genome copy/g of lung.

4. Plasma viral RNA loads (measured using an RT‐qPCR assay) expressed in viral genome copies/ml of plasma.

Data information: Data are represented as mean ± SD (indicated by the error bars). For all graph, each experiment was performed in four biological replicates (N = 4). Exploratory analysis was performed using a two‐way ANOVA with a Sidak’s test correction. Only P‐values ≤ 0.05 were indicated by a * or ^# symbol. For graph a, significant differences between mock‐infected animals and clinical/ISA European/ISA D614 SARS‐CoV‐2‐infected animals were indicated by a *; significant differences between clinical European/ISA D614 SARS‐CoV‐2‐infected animals and ISA European SARS‐CoV‐2 strain‐infected animals were indicated by a ^#. For graph (B), (C), and (D), ** and * symbols indicate significant difference with a P‐value ranging between 0.001–0.01 and 0.01–0.05, respectively (details in Appendix Tables S8 and S9). In other cases, P‐values > 0.05 were considered as not significant and were not displayed on the graph.