Identifying Inhibitors of -1 Programmed Ribosomal Frameshifting in a Broad Spectrum of Coronaviruses

Abstract

Recurrent outbreaks of novel zoonotic coronavirus (CoV) diseases in recent years have highlighted the importance of developing therapeutics with broad-spectrum activity against CoVs. Because all CoVs use -1 programmed ribosomal frameshifting (-1 PRF) to control expression of key viral proteins, the frameshift signal in viral mRNA that stimulates -1 PRF provides a promising potential target for such therapeutics. To test the viability of this strategy, we explored whether small-molecule inhibitors of -1 PRF in SARS-CoV-2 also inhibited -1 PRF in a range of bat CoVs-the most likely source of future zoonoses. Six inhibitors identified in new and previous screens against SARS-CoV-2 were evaluated against the frameshift signals from a panel of representative bat CoVs as well as MERS-CoV. Some drugs had strong activity against subsets of these CoV-derived frameshift signals, while having limited to no effect on -1 PRF caused by frameshift signals from other viruses used as negative controls. Notably, the serine protease inhibitor nafamostat suppressed -1 PRF significantly for multiple CoV-derived frameshift signals. These results suggest it is possible to find small-molecule ligands that inhibit -1 PRF specifically in a broad spectrum of CoVs, establishing frameshift signals as a viable target for developing pan-coronaviral therapeutics.

Keywords: SARS-CoV-2; coronavirus; programmed ribosomal frameshifting; therapeutics; translation.

Figure 1

Pseudoknots from bat coronaviruses. (A) Pseudoknots form when the unpaired bases in an RNA stem loop pair with another single-stranded segment. (B) CoV pseudoknots have a 3-stem architecture, illustrated here for the pseudoknot from SARS-CoV-2. (C) Phylogenetic tree showing five clusters of bat and human CoVs with similar pseudoknot sequences. Representatives from each cluster studied are shown here in cyan for bat CoVs and red for human CoVs.

Figure 2

Efficiency of −1 PRF stimulated by bat CoV frameshift signals. The −1 PRF efficiency measured from cell-free dual-luciferase assays is in the range 25–50% typical of CoVs for all except KY770854 (cluster 3). Error bars represent standard error of the mean from 4–17 replicates.

Figure 3

Drugs modulating −1 PRF in SARS-CoV-2 from screening assays. (A) Results from screening a library of 1814 FDA-approved drugs using a dual-luciferase reporter measured in vitro in rabbit-reticulocyte lysate; most of the hits inhibited −1 PRF (dark grey), but some enhanced it (light grey). Error bars represent the standard error of the mean from 3 replicates. (B) Inhibition of −1 PRF by selected compounds in A549 human lung epithelial cells transfected with a bi-fluorescent reporter system. Error bars represent standard error of the mean from 4 replicates.

Figure 4

Activity of −1 PRF inhibitors against frameshift signals from different CoVs. (A) Change in −1 PRF efficiency compared to basal levels (Figure 2) induced by 20 μM MTDB. Remaining panels show the same for (B) valnemulin, (C) abemaciclib, (D) palbociclib, (E) merafloxacin, and (F) nafamostat. In each case, results for CoVs are shown on left, results for specificity controls on right. Experiments performed in vitro using dual-luciferase reporter in rabbit reticulocyte lysate. Error bars represent standard error of the mean from 3–6 replicates. Insets: chemical structures of inhibitors.