Interfering with nucleotide excision by the coronavirus 3’-to-5’ exoribonuclease

Figures

• 

  Figure 1.. ExoN prefers primed-template-like dsRNA substrates. (A) Purified SARS-CoV-2 nsp14 (60 kDa) and nsp10 (15 kDa) proteins used in this study. Proteins (1 μg) were resolved on a 15% polyacrylamide gel containing SDS and stained with Coomassie. Broad-range molecular weight markers (Mr) and corresponding molecular weights are indicated. (B) Schematic of ssRNA and dsRNA substrates used to measure exoribonuclease activity. RNA sequences are provided in Table 1 and/or Fig. S2. RNAs were labeled on the 5’-end with 32P, also indicated by an asterisk (*). RNAs are designated as primers (P) or templates (T). The numbers indicate the length of the RNA. When a base is designated, this reflects the 3’-terminal nucleotide. (C) Evaluation of ExoN-catalyzed hydrolysis of RNA. Reaction products were resolved by denaturing PAGE and visualized by phosphorimaging. Reactions contained 0.1 μM ExoN (1:1 nsp14:nsp10) and 0.1 μM of the indicated RNA, were incubated at 30 °C for the indicated time, then quenched by addition of EDTA. In both cases, primed-template-like dsRNA substrates were cleaved more efficiently than ssRNA substrates. (D) Exoribonuclease activity is dependent on the nsp14 active site.…

• 

  Figure 2.. Products of nsp15-catalyzed endonucleolytic cleavage are not substrates for ExoN. (A) Schematic of dsRNA substrates used. Both substrates I and II contain 32P-labeled RNAs (*P-3 and *P-2) annealed to the template such that both RNAs have an exposed 3’-end. Substrate III contains a 32P-labeled RNA (*P-2) annealed to the template at a position where an additional RNA (P-3) is annealed downstream and blocks access to the 3’-end of the labeled RNA. Substrate III permits assessment of ExoN cleavage at a nick, as produced by nsp15 endonucleolytic cleavage. (B) ExoN does not initiate hydrolysis at a nick. The 10-nt and 15-nt RNAs are indicated. Reactions contained 0.5 μM ExoN (1:1) and 0.1 μM RNA and were quenched at the indicated times. The 32P-labeled RNAs (*P-3 and *P-2) in substrates I and II that have an exposed 3’-end were efficiently cleaved by ExoN. The 32P-labeled RNA (*P-2) in substrate III was not cleaved. (C-E) ExoN does not hydrolyze termini containing a 3’-phosphate or 2’,3’-cyclic phosphate. The structures of these modifications are shown in panel C. Schematic of dsRNA substrates containing 3’-phosphate and 2’,3’-cyclic phosphate modifications used are shown in panel D.…

• 

  Figure 3.. Monitoring ExoN-catalyzed RNA hydrolysis at single-nucleotide resolution. (A) Schematic of dsRNA substrate used to monitor ExoN activity at single-nucleotide resolution. (B) Reactions contained 0.05 μM ExoN (1:1) and 1 μM RNA or 0.1 μM ExoN (1:1) and 0.05 μM RNA, were incubated for the indicated time, then quenched. Products were resolved using denaturing 20 or 23% polyacrylamide gels. The RNA cleavage products, n-1 to n-8, are indicated. (C,D) Quantitative analysis of kinetics of substrate RNA utilization and/or product RNA formation. Formation of individual products is shown on the right. Panel C reports on the experiment in which RNA was in excess of enzyme; panel D reports on the experiment in which enzyme was in excess of RNA.

• 

  Figure 4.. nsp10 stimulates hydrolysis without affecting processivity or substrate specificity. (A) Effect of nsp10 concentration on the kinetics of nsp14-catalyzed dsRNA hydrolysis. Reactions contained fixed nsp14 (0.1 μM) and varied nsp10 (0 – 2 μM) concentrations. In all cases, nsp14 was pre-mixed with nsp10 on ice 5 min prior to adding to the reaction. Hydrolysis of dsRNA substrate (*P10A-P40:T57) was monitored over a 30 min time course as indicated. Product analysis is shown. (B) Effect of nsp10 concentration on the kinetics of nsp14-catalyzed hydrolysis of ssRNA. Reactions were as in panel A except a ssRNA substrate (*P10A) was used. Product analysis is shown. Products shorter than n-4 are not observed when the 10-nt ssRNA is used.

• 

  Figure 5.. RNA hydrolysis by ExoN is inhibited by a phosphorothioate bond. (A) Structure of phosphorothioate. One of the non-bridging oxygens of the phosphodiester bond is replaced with a sulfur atom, creating Rp and Sp diastereomers. A 50:50 ratio of diastereomers were used in these experiments. (B) Schematic of phosphorothioate-substituted ssRNAs used. (C) Effect of phosphorothioate ExoN-catalyzed hydrolysis of dsRNA. Reactions contained 0.5 μM ExoN (1:1) and 0.1 μM of the indicated RNA, were incubated at the indicated times, then quenched. Product analysis is shown. Fraction of RNA remaining is indicated (% uncleaved). The unmodified RNA is completely degraded; however, ExoN is unable to hydrolyze beyond the phosphorothioate substitution, for half the RNA molecules at least. This observation is consistent with only a single diastereomer being inhibitory.

• 

  Figure 6.. ExoN is inhibited by the Rp diastereomer at the phosphorothioate-substituted scissile bond. (A) The Sp diastereomer of a nucleoside-5’-O-thiotriphosphate is preferentially incorporated by RNA polymerases. Nucleophilic attack at the alpha phosphorous atom leads to inversion of configuration and creation of Rp diastereomer at the phosphorothioate bond. (B) Schematic of assay. Primer extension is initiated by adding an RNA polymerase in the presence of either UTP and CTPαS or ATPαS. CTPαS is a mixture of both the Rp and Sp diastereomers. ATPαS was obtained as the pure Sp diastereomer. Incorporation yields extended products. The last nucleotide to be incorporated is either CMP(αS) or AMP(αS). Once 50–75% of the primers were extended to the end, ExoN was added to the reaction. The reaction was monitored over time for hydrolysis. Unextended primer in reactions served as a useful control to demonstrate the presence of active ExoN in the reaction. (C, D) Analysis of reaction products by denaturing PAGE. Incorporation of CMP and AMP results in removal of the incorporated nucleotide by ExoN. Incorporation of CMPαS and AMPαS (Rp diastereomers) results in a terminated primer that cannot be…

• 

  Figure 7.. Terminal basepair-dependent differences in the kinetics of hydrolysis by ExoN. (A) Schematic of dsRNA substrates used to evaluate the influence of the terminal basepair on kinetics of excision by ExoN. The termini have a A:U, G:U or R:U basepair, where R represents remdesivir. (B) Effect of terminal basepair on hydrolysis by ExoN. Reactions contained 0.05 μM ExoN (1:5) and 1 μM of the indicated dsRNA substrate, were incubated at 30 °C for the indicated times, then quenched. Product analysis is shown. (C) Analysis of the kinetics of substrate utilization and product formation. Here, depletion of substrate and total product formation are monitored, showing no difference in the kinetics. (D) Analysis of the kinetics of formation of each individual product. Cleavage products (n-1 to n-8) were plotted as a function of time for each RNA substrate. The kinetics of formation and utilization of the n-1 product vary between substrates, suggesting that some, but not all steps of the mechanism are agnostic to the sequence and/or complementarity of the terminal basepair.

• 

  Figure 8.. Structure-activity relationships for the ribose of the 3’-terminal nucleotide. (A) Modifications to the ribose of the terminal nucleotide studied here in the context of P10A: 3’-deoxy (-H), 3’-phosphate (-O-PO3), 2’-deoxy (-H), and 2’-O-methyl (-O-CH3); or P10U: 2’-amino (-NH2) and 2’-fluoro (-F). (B,C) Kinetics of cleavage of modified dsRNAs. In panel B, substrate was present in excess of enzyme. In panel C, enzyme was present in excess of substrate. The data were fit to a line (panel B) or to a single exponential (panel C). Reactions contained 0.1 μM ExoN (1:20) and either 1 or 0.05 μM dsRNA substrate and were quenched at the indicated times. The order in which the modified RNAs were cleaved is as follows: unmodified=2’-F > 2’-O-Me > 2’-deoxy > 2’-amino > 3’-deoxy > 3’-phosphate. The kinetics of cleavage using the 3’phosphate RNA is not shown, because cleavage was not detected. The observed rate of cleavage of modified RNAs and fold difference from the unmodified RNA are shown in Table 3.

• 

  Figure 9.. Viperin product (ddhCMP)-terminated RNA is a good substrate for ExoN but 3’-dCMP-terminated RNA is not. (A) Structure of nucleotide analogs used. (B) Schematic of the assay. Primer extension is initiated by adding an RNA polymerase in the presence of UTP and a CTP analog. Incorporation produces n+1 and n+2 products. The last nucleotide to be incorporated is the CTP analog. Once 50–75% of the primers were extended to n+2, ExoN was added to the reaction. The reaction was monitored over time for hydrolysis. Unextended primer in reactions served as a useful control to demonstrate the presence of active ExoN in the reaction. (C) Analysis of reaction products by denaturing PAGE. Only the 3’-dCMP-terminated RNA exhibited a delay in excision. (D) Kinetics of excision of CMP analogs. The quantitation revealed that only 3’-dCMP-terminated RNA exhibited a significant delay in the rate of excision relative to the CMP control, even though ddhCMP also lacks a 3’-OH. Data were fit to a single exponential. Rates are provided in Table 4.

• 

  Figure 10.. Stalled SARS-CoV-2 replication-transcription complexes block access of ExoN to RNA terminus. (A) Schematic of assay. Primer extension was initiated by adding the SARS-CoV-2 replicase-transcription complex (nsp7/nsp8/nsp12) in the presence of ATP and UTP or ATP, UTP, and CTP or a CTP analog. After an incubation time of 60 min, we added ExoN, incubated for the indicated time, and then quenched the reaction. (B) Effect of ExoN on stalled elongation complexes. Incorporation yields n+1, n+2, and n+3 products. When ATP, UTP, and CTP are present, the n+3 product has a properly basepaired 3’-end. The n+4 product is formed by misincorporation and has a mispaired 3’-end. When only ATP and UTP (low concentration, 1 μM UTP) are present, both the n+3 and n+4 products are formed by misincorporation and both products have mispaired 3’-ends. At high concentrations of UTP (100 μM), most of the products observed are at n+3 from misincorporation of U opposite G. ExoN was added to elongation complexes for the indicated time (5 – 60 min). Reaction products were visualized by phosphorimaging after denaturing PAGE. No change in the level of properly paired (n+3) or mispaired (n+4) elongation products…

• 

  Figure 11.. Hypothetical model for ExoN-catalyzed excision. (A) Excision uncoupled to RNA synthesis. The active exoribonuclease, referred to here as ExoN, is a complex of two subunits: nsp10 (regulatory) and nsp14 (catalytic). The concentration of ExoN therefore depends on the value of the dissociation constant for this binding reaction, as well as the concentrations of nsp10 and nsp14. Most studies to date have ignored this equilibrium. Under steady-state conditions in which the RNA substrate (RS) is present in excess of ExoN, RS binds, catalysis occurs, releasing one or two nucleotides, followed by dissociation of the RNA product (RP). Under these conditions, the rate-limiting step is likely release of RP. Modifications to the terminal nucleotide that reduce the rate of hydrolysis to values substantially lower than the rate of RP dissociation should interfere with excision. RP consumption depends on its ability to compete with RS, for example when more than 20% of the substrate is consumed or when ExoN is present in excess of substrate and more than a single turnover occurs. The possibility also exists for product dissociation to be driven by dissociation of the nsp10 and nsp14 subunits.…