The rocaglate CR-31-B (-) inhibits SARS-CoV-2 replication at non-cytotoxic, low nanomolar concentrations in vitro and ex vivo

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the causative agent of COVID-19, a severe respiratory disease with varying clinical presentations and outcomes, and responsible for a major pandemic that started in early 2020. With no vaccines or effective antiviral treatments available, the quest for novel therapeutic solutions remains an urgent priority. Rocaglates, a class of plant-derived cyclopenta[b]benzofurans, exhibit broad-spectrum antiviral activity against multiple RNA viruses including coronaviruses. Specifically, rocaglates inhibit eukaryotic initiation factor 4A (eIF4A)-dependent mRNA translation initiation, resulting in strongly reduced viral RNA translation. Here, we assessed the antiviral activity of the synthetic rocaglate CR-31-B (-) against SARS-CoV-2 using both in vitro and ex vivo cell culture models. In Vero E6 cells, CR-31-B (-) inhibited SARS-CoV-2 replication with an EC₅₀ of ~1.8 nM. In primary human airway epithelial cells, CR-31-B (-) reduced viral titers to undetectable levels at a concentration of 100 nM. Reduced virus reproduction was accompanied by substantially reduced viral protein accumulation and replication/transcription complex formation. The data reveal a potent anti-SARS-CoV-2 activity by CR-31-B (-), corroborating previous results obtained for other coronaviruses and supporting the idea that rocaglates may be used in first-line antiviral intervention strategies against novel and emerging RNA virus outbreaks.

Keywords: Antiviral activity; COVID-19; Rocaglate; SARS-CoV-2; Translation initiation; eIF4A.

Graphical abstract

Fig. 1

Effect of CR-31-B (−) on dual reporter gene expression from constructs containing different 5′-UTRs. (A) Structure of the synthetic rocaglate CR-31-B (−). (B) Schematic illustration of the dual luciferase reporter construct used to determine eIF4A-dependent translation of coronavirus 5′-UTRs. (C) Effects of 5 and 10 nM CR-31-B (−) on reporter gene expression in the context of 5′-UTRs from three human coronaviruses, HCoV-229E, MERS-CoV and SARS-CoV-2. The 5′-UTR of the human β-globin mRNA and the unstructured (AC)₁₅ sequence served as negative controls, while the (AG)₁₅ polypurine sequence served as a positive control. Experiments were performed using previously described protocols (Müller et al., 2018a, 2020) with at least three independent biological replicates.

Fig. 2

Dose-dependent antiviral activity of the synthetic rocaglate CR-31-B (−) in SARS-CoV-2 infected Vero E6 cells. (A) Vero E6 cells were treated for 24 h with the indicated CR-31-B (−) concentrations. Cell viability (compared to that of untreated cells) was determined by MTT assay (n = 8) as described previously (Müller et al., 2018a). (B) SARS-CoV-2 titers in supernatants collected from infected Vero E6 cells (MOI = 0.1) treated with the indicated CR-31-B (−) concentrations were collected at 24 h p.i. (n = 6) and virus titers were determined by plaque assay. Significance levels compared to the results for untreated cells were determined by the two-tailed Mann Whitney U test and are indicated as follows: *, P < 0.05; **, P < 0.005. Data from six independent experiments were used to calculate the EC₅₀ value by non-linear regression analysis. (C) Representative Western blot analysis of SARS-CoV-2 N protein accumulation (top panel) after treatment with CR-31-B (−) and CR-31-B (+). Vero E6 cells were infected with SARS-CoV-2 (MOI = 1) or left uninfected and treated with the indicated CR-31-B concentrations for 24 h p.i. Protein accumulation was analyzed by Western blotting using polyclonal rabbit anti-SARS nucleocapsid protein antibody (Rockland) and mouse-anti actin antibody (abcam), respectively, each diluted 1:500 in PBS containing 1% bovine serum albumin (BSA). Beta-actin (lower panel) was used as a loading control (n = 3). (D) Representative immunofluorescence analysis to determine the effects of CR-31-B (+) and CR-31-B (−) on viral dsRNA accumulation in SARS-CoV-2-infected Vero E6 cells. Cells were infected (MOI = 1) and cultivated in medium containing the indicated CR-31-B concentrations. Cells were fixed at 24 h p.i. and analyzed by confocal microscopy using a mouse anti-dsRNA mAb (J2, SCICONS English & Scientific Consulting Kft, red) that detects a viral RNA replication intermediate (red) (Müller et al., 2018b). Cell nuclei were stained with DAPI (blue) (n = 2). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

Comparison of antiviral effects of CR-31-B (−) and silvestrol using differentiated NHBE cells infected with SARS-CoV-2. (A) NHBE cells (Lonza, CC-2540) obtained from a 13-year-old Caucasian boy (Donor 1) and a 36-year-old Caucasian man (Donor 2), which both were non-smoking and lacking respiratory pathology, were seeded on collagen IV-coated transwell plates (Corning Costar, CLS3470-48 EA) and grown in a mixture of DMEM (Invitrogen) and BEGM (Lonza; CC-4175) supplemented with retinoic acid (75 nM). After reaching confluence, the cells were cultivated under air-liquid conditions for at least four additional weeks to allow differentiation into pseudostratified human airway epithelia. (B and C) Differentiated NHBE cells were infected with SARS-CoV-2 (MOI = 3) and treated with silvestrol (B), CR-31-B (−) or CR-31-B (+) (C) at the indicated concentrations. At the indicated time points p.i., the apical surface of the cells was incubated with 100 μl PBS for 15 min and SARS-CoV-2 titers in the cell culture supernatants were determined by virus plaque assay.