Biochemical characterization of a recombinant SARS coronavirus nsp12 RNA-dependent RNA polymerase capable of copying viral RNA templates

Abstract

The severe acute respiratory syndrome coronavirus (SARS-CoV) RNA genome is replicated by a virus-encoded RNA replicase, the key component of which is the nonstructural protein 12 (nsp12). In this report, we describe the biochemical properties of a full-length recombinant SARS-CoV nsp12 RNA-dependent RNA polymerase (RdRp) capable of copying viral RNA templates. The purified SARS-CoV nsp12 showed both primer-dependent and primer-independent RNA synthesis activities using homopolymeric RNA templates. The RdRp activity was strictly dependent on Mn(2+). The nsp12 preferentially copied homopolymeric pyrimidine RNA templates in the absence of an added oligonucleotide primer. It was also able to initiate de novo RNA synthesis from the 3'-ends of both the plus- and minus-strand genome of SARS-CoV, using the 3'-terminal 36- and 37-nt RNA, respectively. The in vitro RdRp assay system established with a full-length nsp12 will be useful for understanding the mechanisms of coronavirus replication and for the development of anti-SARS-CoV agents.

Fig. 1

Purification of full-length SARS-CoV nsp12 expressed in E. coli. (A) Schematic diagram of a full-length SARS-CoV nsp12. The unique domain, the catalytic domain, and the conserved RdRp active-site residues of the nsp12 are indicated. The deduced N-terminal sequences of the indicated nsp12 protein are shown. (B-D) The nsp12 protein was expressed in E. coli and purified by affinity chromatography using an Ni-NTA agarose column (B), anion-exchange chromatography using a Q-Sepharose column (C), and GFC using a Superdex 200 column (D). SARS-CoV nsp12 bound to an Ni-NTA agarose and a Q-Sepharose column were eluted by increasing concentrations of imidazole and NaCl, respectively. Eluted proteins and fractions collected from the Superdex 200 column were separated by 10 % SDS-PAGE and stained with Coomassie blue. Arrowheads indicate the position of SARS-CoV nsp12. In panel (D), proteins identified by MS/MS analysis are shown. (E) The final purified recombinant nsp12 (2 μg) was resolved by SDS-PAGE and visualized by Coomassie blue staining

Fig. 2

Primer- and Mn ²⁺ -dependent RNA synthesis by nsp12. RdRp assays were performed with a homopolymeric poly(A) RNA template in the presence (+) or absence (−) of the 20-nt complementary ribonucleotide primer U₂₀ (A and B, as indicated; D-G, in the presence of U₂₀) or with the viral-genome-derived RNA templates (C, see schematic diagrams of the RNA templates in Fig. 4 A and D). The reactions were carried out in the presence of 2 mM Mn²⁺ (A and E-G), 2 mM Mg²⁺ or Mn²⁺ (B-C), or increasing concentrations of Mn²⁺ (D). Radioisotope-labeled RNA products were resolved on an 8 M urea-5 % polyacrylamide gel before drying and being exposed to x-ray film for autoradiography. In (D-G), the relative RdRp activity of nsp12 is presented as the percentage of that obtained under each optimal condition

Fig. 3

Primer dependence of nsp12 in RNA synthesis initiation from homopolymeric RNA templates. (A) RdRp assays were performed using the indicated homopolymeric RNA templates in the absence (−) or presence (+) of 20-mer complementary ribonucleotide primer. Mn²⁺ (2 mM) was used as the metal cofactor. The amount of ³²P-GMP incorporated in the reaction with the poly(C)/(rG)₂₀ template was measured to be ~5 × 10⁴ cpm. (B) RdRp assays were performed with a poly(C) template in the presence of Mg²⁺ or Mn²⁺. (C) RdRp activities of wild-type nsp12, a mutant nsp12 carrying a SDD-to-SAA substitution [nsp12(SAA)], and the unique domain of nsp12 (nsp12-unique domain) were tested with a poly(C) template in the absence of the (rG)₂₀ primer. (D) RdRp assays with a poly(C) template were carried out in the absence (−) or presence of actinomycin D (20 μg/ml) or rifampicin (20 μg/ml). RNA products were analyzed as in Fig. 2

Fig. 4

Mapping of minimal RNA domains required for initiation of plus-strand and minus-strand viral RNA synthesis. (A) Schematic representation of 3’-UTR RNA templates without or with various lengths (13-32 nt) of polyadenylate and deletion derivatives of the 3’-UTR lacking a poly(A) tail (3’-UTR339∆A). The most stable secondary structure of 3’-UTR36∆A predicted by the Mfold program [40] is shown. (B) RdRp assays were performed with the indicated RNA templates. The numbers following the +sign of the series of 3’-UTR339∆A RNA templates indicate the length of the poly(A) tail. Radioisotope-labeled RNA products were resolved on a medium-size 8 M urea-8 % acrylamide gel (20 cm × 20 cm) and subjected to autoradiography. (C) Minimal cis-acting RNA elements required for minus-strand RNA synthesis were mapped by RdRp assays using the indicated deletion derivatives of the 3’-UTR36∆A. The numbers following the letters “3’-UTR” indicate the corresponding size of each template. RNA products were analyzed as in (B). Template positions are indicated with arrowheads. (D) Deletion derivatives of c5’-UTR RNA templates are depicted on schematic diagrams. The most stable secondary structure of the c5’-UTR 38-nt RNA template predicted by the Mfold program is shown. The numbers following the letters “c5’-UTR” represent the corresponding size of each template. (E) Minimal cis-acting RNA elements required for plus-strand RNA synthesis were mapped by RdRp assays using the indicated RNA templates. Template positions are indicated by arrowheads