The short isoform of the host antiviral protein ZAP acts as an inhibitor of SARS-CoV-2 programmed ribosomal frameshifting

Figures

• 

  Fig. 1. In vitro RNA-antisense purification-based discovery of protein interactors of the SARS-CoV-2 –1PRF element. A Schematic representation of the relevant genomic segment of SARS-CoV-2 as well as the location of the –1PRF element. B Schematic of in vitro interactome capture of protein interactors of the SARS-CoV-2 –1PRF element. In vitro synthesized RNA fragment numbered 1–84 corresponding to nucleotides 13456–13570 of the SARS-CoV-2 genome, was incubated with lysates of naïve HEK293 cells as well as SARS-CoV-2-infected and uninfected Calu-3 cells. The –1PRF RNA was captured by a biotinylated antisense DNA oligo and isolated proteins were subjected to LC-MS/MS. C Representative scatter plot of log2-ratios comparing proteins captured in uninfected vs. SARS-CoV-2-infected Calu-3 cells. Core interactors common between uninfected and SARS-CoV-2-infected Calu-3 cells as well as uninfected HEK293 cells are highlighted in blue, ZAP is highlighted in pink. D Heatmap representing the enrichment (log2) of core interactors. See also Supplementary Fig. 1D.

• 

  Fig. 2. A functional screen of SARS-CoV-2 –1PRF element interactors. A Schematic representation of the dual-fluorescence frameshift reporter construct. EGFP and mCherry are separated by a self-cleaving 2 A peptide as well as by a stop codon in-frame with EGFP. As a result, 0-frame translation would produce only EGFP, whereas –1PRF would produce both EGFP and mCherry. The ratio of mCherry to EGFP fluorescence is used to quantify the FE. The trans-factor construct is an N-terminal fusion of ECFP with the protein of interest to be analyzed. The control construct consists of ECFP alone. B Confocal microscopy images of cells transfected with the EGFP-mCherry control (CC- no –1PRF site included after EGFP and mCherry in-frame with EGFP), –1PRF, and no PRF (no –1PRF site and stop codon after EGFP) constructs. The size bar represents 50 µm. n = 1 independent experiment. C Comparison of relative FE of cells overexpressing trans-factors as ECFP fusion proteins. Data points represent the mean ± s.d. (n = 3 independent experiments). P values were calculated using an ordinary unpaired one-sided ANOVA comparing every condition to the ECFP control. ZAP-L and ZAP-S were separately compared to each…

• 

  Fig. 3. Effect of zinc-finger antiviral protein (ZAP) on 1a/1b –1 frameshifting in vitro. A The strategy of the in vitro translation assay using rabbit reticulocyte lysate (RRL) and the experimental workflow to study ribosome association of ZAP-S. B Schematics of the N-terminal FLAG-tagged frameshifting reporter consisting of the nucleotides 12686-14190 (~1.5 kb) of the SARS-CoV-2 genome. RNAs were translated in RRL in the presence of increasing concentrations of ZAP-S ranging from 0 to 3 µM. FLAG-tagged peptides generated by ribosomes that do not frameshift (no –1PRF) or that enter the −1 reading frame (−1PRF) were identified via western blotting using anti-DDDDK antibody. FE was calculated as previously described11, by the formula: Intensity (–1-frame)/ (Intensity (–1-frame) + Intensity (0-frame)). Size markers - M (Marker), –1PRF M (–1-frame marker), and no –1PRF M (0-frame marker). n = 3 independent experiments. C Changes in FE observed in the presence of ZAP-S from (B) (normalized to 0 µM ZAP as shown in B). P values were calculated using an ordinary unpaired one-sided ANOVA comparing every concentration to the no ZAP control. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.…

• 

  Fig. 4. In vitro characterization of ZAP-S interaction with SARS-CoV-2 –1 PRF RNA. A Proposed structure of the PRF element of SARS-CoV-2. Nucleotide substitutions in the compensatory mutant are indicated (arrowheads). B Schematic representations of the RNAs studied. C–H Microscale thermophoresis assay to monitor ZAP-S binding to (C) Full PRF, (D) ΔSL2 mutant, (E) ΔSL3 mutant, (F) ΔSL2 + 3 mutant, (G) compensatory mutant, (H) scrambled mutant. Unlabeled protein (40 pM–2 µM) was titrated against 3′ pCp-Cy5 labeled RNA (5 nM) and thermophoresis was recorded at 25 °C with 5% LED intensity and medium MST power. Change in fluorescence (ΔFnorm) was measured at MST on-time of 2.5 s. Data were analyzed for ΔFnorm using standard functions of MO. Affinity Analysis software and data was plotted and KD was determined using Graphpad Prism 9.2.0. Data represent mean ± s.d. of three measurements (n = 3). For the related thermophoretic traces, see also Supplementary Fig. 4A-F. For the related DNA sequences of the mutants, see also Supplementary Table 2.

• 

  Fig. 5. Single molecule characterization of mechanical properties of SARS-CoV-2 PRF RNA in the presence of ZAP-S. A Schematic illustrating optical tweezers experiments. RNA was hybridized to single-stranded DNA handles flanking the SARS-CoV-2 frameshift site and conjugated to functionalized beads. A focused laser beam was used to exert pulling force from one end of the molecule. The force was gradually increased until the RNA was fully unfolded (bottom). B 3D structure of SARS-CoV-2 pseudoknot RNA (PK) derived from51 and colored according to the scheme used in Fig. 4. C Schematic representations of the RNAs studied. D–I Example unfolding and refolding traces of PK in the presence or absence of ZAP-S, “F” denotes the folded state, “I” the intermediate, and “U” the fully unfolded state, (D) PK (N = 273 FD curves from 24 molecules no ZAP-S, N = 219 FD curves from 24 molecules +ZAP-S samples), (E) ΔSL2 mutant (N = 146 FD curves from eight molecules no ZAP-S, N = 122 FD curves from eight molecules +ZAP-S samples), (F) ΔSL3 mutant (N = 127 FD curves from 12 molecules no ZAP-S, N = 163 FD curves from 11 molecules +ZAP-S samples), (G) ΔSL2 + 3 mutant (N = 216 FD curves from eight molecules no…

• 

  Fig. 6. Model for ZAP-S mediated inhibition of SARS-CoV-2 frameshifting. 1 Upon infection, the viral RNA is translated by the cellular machinery, and 40% of translation events yield the 1a/1b polyprotein through –1PRF. 2 Infection also leads to the induction of antiviral factors including ZAP-S. 3 ZAP-S binding to the frameshift RNA alters RNA refolding and thereby reduces the chance of elongating ribosomes to encounter the stimulatory structure. Thus, the elongation pause is too short for codon-anti-codon interactions to be established in the –1-frame and ZAP-S allows translation to proceed without a strong roadblock effect. This leads to termination at the canonical 0-frame UAA stop codon found just downstream of the slippery sequence. The resulting decrease in the amounts of the 1a/1b polypeptides reduces the levels of the viral RNA-dependent RNA polymerase (RdRP) from the –1-frame.