Functional screen reveals SARS coronavirus nonstructural protein nsp14 as a novel cap N7 methyltransferase

Abstract

The N7-methylguanosine (m7G) cap is the defining structural feature of eukaryotic mRNAs. Most eukaryotic viruses that replicate in the cytoplasm, including coronaviruses, have evolved strategies to cap their RNAs. In this report, we used a yeast genetic system to functionally screen for the cap-forming enzymes encoded by severe acute respiratory syndrome (SARS) coronavirus and identified the nonstructural protein (nsp) 14 of SARS coronavirus as a (guanine-N7)-methyltransferase (N7-MTase) in vivo in yeast cells and in vitro using purified enzymes and RNA substrates. Interestingly, coronavirus nsp14 was previously characterized as a 3'-to-5' exoribonuclease, and by mutational analysis, we mapped the N7-MTase domain to the carboxy-terminal part of nsp14 that shows features conserved with cellular N7-MTase in structure-based sequence alignment. The exoribonuclease active site was dispensable but the exoribonuclease domain was required for N7-MTase activity. Such combination of the 2 functional domains in coronavirus nsp14 suggests that it may represent a novel form of RNA-processing enzymes. Mutational analysis in a replicon system showed that the N7-MTase activity was important for SARS virus replication/transcription and can thus be used as an attractive drug target to develop antivirals for control of coronaviruses including the deadly SARS virus. Furthermore, the observation that the N7-MTase of RNA life could function in lieu of that in DNA life provides interesting evolutionary insight and practical possibilities in antiviral drug screening.

Fig. 1.

Functional screening for RNA cap-forming functions encoded by SARS-CoV. (A) YBS40 could be complemented by SARS-CoV-nsp14 and the positive control pYX232-HCM1 but not by other nonstructural proteins of SARS-CoV. (B) The nsp14 of both SARS-CoV and TGEV could rescue the growth of YBS40 on 5-FOA medium. The known N7-MTases including nsP1 of SFV and NS5-MTase domain of WNV could not substitute for the yeast N7-MTase activity.

Fig. 2.

Biochemical analyses of the guanine-N7 methyltransferase activity of nsp14 of SARS-CoV. (A) Purification and SDS/PAGE analysis of recombinant nsp14 (lane 2), point mutants (lanes 3–6), truncation mutants of nsp14 (lanes 7–8), nsp3-SUD domain (lane 9), nsp12N (lane 10), and nsp16 (lane 11) of SARS-CoV. nsp12N represents a distinct N-terminal domain without predicted function in nsp12. The sizes of protein markers (lane 1) are indicated on the Left. (B) TLC analysis of nuclease P1-resistant cap structures released from the G*pppA-RNA treated by nsp14, nsp3-SUD, nsp12N, and nsp16, respectively (lanes 3–6). The asterisk indicates that the following phosphate is ³²P-labeled. The positions of origin and migration of m7G*pppA/G*pppA (lanes 1–2) are indicated on the Left. The positive controls for m7G*pppA and G*pppA were generated by vaccinia capping enzyme. (C) TLC analysis of tobacco acid pyrophosphatase-digested products from the reactions in panel B. The positions of origin, m7G*p and G*p are indicated on the Left. (D) TLC analysis of nuclease P1-resistant cap structures released from the G*pppG-RNA treated by nsp14, nsp14ΔN61, nsp14ΔC17, mutant D331A, nsp14# (stored in elution buffer, otherwise identical to nsp14 in lane 3), nsp3-SUD, nsp16, respectively (lanes 3–9). The positions of origin and migration of m7G*pppG/G*pppG (lanes 1–2) are indicated on the Left.

Fig. 3.

The domain structure of SARS-CoV nsp14 and its mutants. The core domains of ExoN and MTase are indicated with open box and solid black box, respectively. The conserved residues DxG (amino acid 331–333) in the MTase domain are putative SAM-binding sites. The conserved residues D-E-D-H in the ExoN domain are indicated. The truncation mutants are named as the number of the deleted amino acids from either N or C terminus. The capability for complementing yeast growth and the activities of MTase and ExoN are presented in the Right panel. The values corresponding to + +, + and − are >60% activity of wild-type protein (or wild-type growth), >15% (or slow growth), and <15% (or no growth). ND, not determined.

Fig. 4.

MTase activity analysis of truncation mutants of SARS-CoV nsp14. (A) YBS40 (ade2) transformed with the N-terminal truncation mutants of SARS-CoV nsp14. ΔN31 and N61 could readily rescue the cell growth of YBS40 on 5-FOA medium while cells rescued by ΔN78 grew slower than that with ΔN31 and ΔN61 as shown by lesser red pigment accumulation and smaller cell density. In contrast, ΔN90 and other N-terminal truncation lost the MTase activity in yeast. (B) YBS40 transformed with the C-terminal truncation mutants of SARS-CoV nsp14. The mutant ΔC6 (amino acid 1–521) could rescue YBS40 on 5-FOA medium but the cells grew obviously slower than those with wild-type nsp14. All other C-terminal truncations abolished the MTase activity of nsp14 in yeast cells. (C) TLC analysis of nuclease P1-resistant cap structures released from the G*pppA-RNA methylated by nsp14, nsp14ΔN61, nsp14ΔC17, nsp12N, respectively (lanes 3–6). The positions of origin and migration of m7G*pppA/G*pppA (lanes 1–2) are indicated on the Left.

Fig. 5.

Effects of MTase and ExoN active-site mutations on enzymatic activities of nsp14 and on SARS-CoV replicon. (A) YBS40 was transformed with the point mutants of SARS-CoV nsp14. D90A/E92A, D243A, and H268L mutations in the ExoN domain could rescue the cell growth of YBS40 on 5-FOA medium while D331A in the predicted MTase domain lost the MTase activity in yeast cells. (B) TLC analysis of nuclease P1-resistant cap structures released from the G*pppA-RNA treated by wild-type nsp14 and its mutants D90A/E92A, D243A, H268L, and D331A (lanes 3–7). The nsp3-SUD and nsp16 (lanes 8–9) were used as negative controls. The positions of origin and migration of m7G*pppA/G*pppA (lanes 1–2) are indicated on the Left. (C) Analysis of 3′-to-5′ ExoN activity of nsp14 and its mutants D90A/E92A, D243A, H268L, and D331A (lanes 5–9) in 10% urea-PAGE using single-stranded RNA substrates that were ³²P-labeled at the 5′-end. The mock treatment or with nsp3-SUD, nsp12N, and nsp16 were used as negative controls (lanes 1–4). (D) Quantitation of ExoN activity by analyzing the band intensity shown in Fig. 5C with ImageQuant software. The values in the y axis indicate the ratios of digestion products to input RNA within each lane. All products migrating below full-length substrate were quantitated together as digestion products. Input RNA represents digestion products plus undigested full-length substrate. (E) Effect of D331A mutation on luciferase reporter of SARS-CoV replicon. The reporter-carrying replicon (rep) and a deficient deletion mutant (ΔMluI) were described previously (22). (F) Quantitative PCR of subgenomic RNAs with the same set of samples as Fig. 5E (Left y axis). β-Actin mRNA was detected as internal control to show that equal amount of RNA was used (Right y axis). Graphs show the mean of RNA copy number of 3 experiments ± SD.