Analysis of multimerization of the SARS coronavirus nucleocapsid protein

Abstract

Severe Acute Respiratory Syndrome (SARS), an emerging disease characterized by atypical pneumonia, has recently been attributed to a novel coronavirus. The genome of SARS Coronavirus (SARS-CoV) has recently been sequenced, and a number of genes identified, including that of the nucleocapsid protein (N). It is noted, however, that the N protein of SARS-CoV (SARS-CoV N) shares little homology with nucleocapsid proteins of other members of the coronavirus family [Science 300 (2003) 1399; Science 300 (2003) 1394]. N proteins of other coronavirus have been reported to be involved in forming the viral core and also in the packaging and transcription of the viral RNA. As data generated from some viral systems other than coronaviruses suggested that viral N-N self-interactions may be necessary for subsequent formation of the nucleocapsid and assembly of the viral particles, we decided to investigate SARS-CoV N-N interaction. By using mammalian two-hybrid system and sucrose gradient fractionations, a homotypic interaction of N, but not M, was detected by the two-hybrid analysis. The mammalian two-hybrid assay revealed an approximately 50-fold increase in SEAP activity (measurement of protein-protein interaction) in N-N interaction compared to that observed in either M-M or mock transfection. Furthermore, mutational analyses characterized that a serine/arginine-rich motif (SSRSSSRSRGNSR) between amino acids 184 and 196 is crucial for N protein oligomerization, since deletion of this region completely abolished the N protein self-multimerization. Finally, the full-length nucleocapsid protein expressed and purified from baculovirus system was found to form different levels of higher order structures as detected by Western blot analysis of the fractionated proteins. Collectively, these results may aid us in elucidating the mechanism pertaining to formation of viral nucleocapsid core, and designing molecular approaches to intervene SARS-CoV replication.

Fig. 1

Yeast two-hybrid assay for N protein self-interaction. S. cerevisiae AH109 cells (1.5 ml of overnight culture) were transformed with 0.1 μg pGBK-N and pGAD-N plasmid DNA using the polyethylene glycol/lithium acetate method and plated on SD/LEU/−TRP and SD/−ADE−HIS/−LEU/−TRP drop-out plates and grown at 30 °C for 4 days. pGBK/T7 + pGAD/T7 and pGBK-M + pGAD-M were used as negative controls. Growth of the transformants on the SD/LEU/−TRP plate indicates that both pM- and pVP-related plasmids have been delivered into yeast cells, while the growth on SD/−ADE−HIS/−LEU/−TRP plate reveals activation of reporter genes, and thus interactions of proteins.

Fig. 2

Mammalian two-hybrid analysis on N protein self-interaction. Vero cells were co-transfected with full-length SARS-CoV N protein constructs, pM-N + pVP-N along with pG5SEAP reporter vector, and incubated for 48 h at 37 °C in a CO₂ incubator. Co-transfection samples, the cloning vector pM + pVP16, and SARS-CoV M protein constructs pM-M + pVP-M, were designated as negative controls. Error bars represent standard deviation (usually less than 10%).

Fig. 3

Western blot analysis of sucrose gradient fractionated His₆-tagged N protein. N protein was cloned into the baculovirus pBlueBacHis2 vector and transfected and expressed in Sf21 cells as described in Materials and methods. Roughly 5 × 10⁶ cells were lysed and analyzed by sucrose gradient centrifugation followed by Western blotting detection of the viral proteins. The control sample is a baculovirus construct with pBlueBAcHis2 alone.

Fig. 4

(A) Schematic description of N gene sequential deletion. N gene was truncated from 5^′, primer sequences are indicated in Table 1. (B) Mammalian two-hybrid analysis of truncation mutant constructs of N protein in comparison with full-length N protein. pM-N mutants were co-transfected with pVP16-N and the reporter vector pG5SEAP into Vero cells. Chemiluminescent activity of secreted alkaline phosphatase present in cell culture medium was assayed as described in Materials and methods. Data represent an increase in chemiluminescence as compared to co-transfected vector controls. Error bars represent standard deviation (usually less than 10%). (C) Mammalian two-hybrid analysis of the requirement of the serine/arginine-rich region for the N protein self-interaction. The co-transfected samples were full-length N protein constructs pM-N and pVP-N, serine/arginine-rich region deletion mutant ΔpM-N/SR and pVP-N, and SARS-CoV M protein pM-M and pVP-M. The reporter vector pG5SEAP was also co-transfected in each group. Chemiluminescent activity of secreted alkaline phosphatase in cell culture medium was analyzed as described in Materials and methods. Error bars represent standard deviation (usually less than 10%).

Fig. 5

Co-localization analysis of SARS-CoV N protein. Constructs of pAsRed-N and ΔpAsRed-N/SR were co-transfected with the nuclear localization vector pECFP-nuc, respectively. Thirty hours post-transfection, cells were visualized with a fluorescent microscope. From A1 to A3 are the N protein, pECFP-nuc, and the super imposed image of A1 and A2. From B1 to B3 are the SR-deletion mutant, pECFP-nuc, and the superimposed image of B1 and B2.