VPg-primed RNA synthesis of norovirus RNA-dependent RNA polymerases by using a novel cell-based assay

Abstract

Molecular studies of human noroviruses (NoV) have been hampered by the lack of a permissive cell culture system. We have developed a sensitive and reliable mammalian cell-based assay for the human NoV GII.4 strain RNA-dependent RNA polymerase (RdRp). The assay is based on the finding that RNAs synthesized by transiently expressed RdRp can stimulate retinoic acid-inducible gene I (RIG-I)-dependent reporter luciferase production via the beta interferon promoter. Comparable activities were observed for the murine norovirus (MNV) RdRp. RdRps with mutations at divalent metal ion binding residues did not activate RIG-I signaling. Furthermore, both NoV and MNV RdRp activities were stimulated by the coexpression of their respective VPg proteins, while mutations in the putative site of nucleotide linkage on VPg abolished most of their stimulatory effects. Sequencing of the RNAs linked to VPg revealed that the cellular trans-Golgi network protein 2 (TGOLN2) mRNA was the template for VPg-primed RNA synthesis. Small interfering RNA knockdown of RNase L abolished the enhancement of signaling that occurred in the presence of VPg. Finally, the coexpression of each of the other NoV proteins revealed that p48 (also known as NS1-2) and VP1 enhanced and that VP2 reduced the RdRp activity. The assay should be useful for the dissection of the requirements for NoV RNA synthesis as well as the identification of inhibitors of the NoV RdRp.

Fig. 1.

RNA synthesis-competent NoV RdRp can activate innate immune receptor-mediated signaling. (A) Typical genome organization of human NoV. (B) Western blot of the expression of wild-type or an active site mutant NoV RdRp (norovirus Hu/GII.4/MD-2004/2004/US). Cells transfected to express the FLAG-tagged WT (NoV Rp) or the active site mutant (Rp^GAA) of the NoV GII.4 RdRp were lysed in sample buffer, resolved with a 4-to-12% bis-Tris gel, and subjected to Western blotting with mouse monoclonal antibody to FLAG tag (Sigma). HRP-coupled rabbit anti-mouse secondary antibody was used to detect the primary antibody, and the signal was developed with the ECL Plus Western blot detection system (Amersham, United Kingdom). The sizes of the relevant molecular weight markers are shown to the left of the Western blot image. (C) Results from the NoV-5BR assay for the NoV RdRp in HEK 293T cells. Plasmids that expressed the proteins are denoted on the x axis. 1b5B is the RdRp of HCV genotype 1b. Ratios of firefly to Renilla luciferase activities are shown on the y axis. The data are the means from three replicates and are representative of three independent experiments, and the standard errors are shown above the bars. (D) Results from a NoV-5BR assay performed in cultured human hepatocytes (Huh-7). The data are the means from three replicates, and the standard errors are shown above the bars. (E) Effects of Mn²⁺ or Mg²⁺ on NoV RdRp activity. The divalent metals were added directly to the medium of the cultured HEK 293T cells. The control reaction monitored the effect of Mn²⁺ on cells that expressed the RIG-I receptor and the two luciferases but not the GII.4 polymerase.

Fig. 2.

MDA5, but not TLR3, can replace RIG-I in the NoV-5BR assay. (A) The NoV RdRp can signal through MDA5 in 293T cells. 1b5B is the RdRp from HCV. Rp^GAA is an active site mutant of the NoV RdRp. (B) TLR3 signaling was not activated by the products of the NoV RdRp. A TLR3 mutant defective in ligand binding, H539E, served as a control to illustrate signaling that was dependent on the WT TLR3. pI:C denotes poly(I:C), a TLR3 agonist, which was added to the medium of cells to a final concentration of 500 ng/ml.

Fig. 3.

RNA synthesis-competent MNV RdRp can activate innate immune receptor-mediated signaling. (A) Western blot of the FLAG-tagged MNV proteins produced in HEK 293T cells. A monoclonal antibody recognizing the FLAG epitope was used to detect the MNV RdRp and VPg proteins. HRP-coupled rabbit anti-mouse secondary antibody was used to detect the primary antibody, and the signal was developed with the ECL Plus Western blot detection system (Amersham, United Kingdom). (B) WT MNV RdRp can induce RIG-I signaling in HEK 293T cells. Rp^GAA is the active site mutant of MNV RdRp. (C) Sequence alignment of VPg proteins from representative members of the Caliciviridae. The predicted site of RNA linkage in the calicivirus VPg is highlighted in bold. Tyr26 is the site of nucleotidylylation in MNV VPg, while the site in NoV corresponds to Tyr27. Tyr117 of the MNV VPg protein, reported to be important for nucleotidylylation in vitro, is also highlighted in bold. The conserved signature motif of the calicivirus VPg proteins is shown (E/DEYDEΩ). (D) Effects of coexpressed wild-type or mutant VPgs on the activity of the MNV polymerase. The presence of a plasmid used in the transfection of HEK 293T cells is shown by a plus symbol below the x axis. A plus symbol in the row titled “Vec” denotes that an empty vector was transfected into the cells. The data are the means from three replicates and are representative of three independent experiments, and the standard errors are shown above the bars. P values less than 0.05 were considered statistically significant.

Fig. 4.

Y26 of the MNV VPg is required for successful MNV infection. (A) Virus yield following fowlpox virus-mediated reverse genetics recovery from cDNA constructs of either wild-type MNV (Rp), RdRp active site mutant (Rp^GAA), or constructs containing the mutations Y26F or Y117F in VPg. Baby hamster kidney cells previously infected with fowlpox virus expressing T7 RNA polymerase were transfected with each of the cDNA clones. At 24 h posttransfection the cultures were freeze-thawed, and the virus yields were determined in permissive RAW 264.7 cells. (B) Western blot of MNV proteins from BHK21 cells. One microgram of each MNV cDNA clone was transfected using Lipofectamine 2000 (Invitrogen) into the cells infected with fowlpox virus expressing T7 RNA polymerase. At 24 h posttransfection samples were separated by SDS-PAGE, transferred to nitrocellulose, and probed with rabbit polyclonal anti-NS7. Antibody binding was detected with goat anti-rabbit DyLight800 conjugate (Thermo Scientific) before results were read with a Licor Odyssey apparatus. (C) Virus titer and plaque phenotype of the viruses recovered in the experiment described for panel A.

Fig. 5.

Effects of coexpressed VPg on NoV GII.4 RdRp activity. (A) Western blot showing that the NoV VPg WT and the Y27A mutant proteins are expressed to comparable levels. (B) Effects of NoV and MNV VPgs on homologous and heterologous combinations of the NoV or MNV RdRps. W, WT VPg; M, mutant VPg. A plus symbol below the x axis shows the presence of the respective plasmids used in the transfection of the HEK 293T cells. The data are the means from three replicates and are representative of three independent experiments, and the standard errors are shown above the bars. Statistical analysis was performed by using the pairwise Student t test, and the P values are shown above the samples that were compared. (C) The NoV VPg does not enhance the activities of the 1b HCV RdRp or the BMV 2a RdRp in transiently transfected HEK 293T cells. The data are the means from three replicates and are representative of two independent experiments, and the standard errors are shown above the bars. Statistical analysis was performed by using the pairwise Student t test. The difference between the NoV Rp and NoV Rp+VPg was statistically significant (P < 0.05).

Fig. 6.

VPg-primed RNA synthesis in the presence of NoV RdRp. (A) Western blot analysis of VPg immunoprecipitated from HEK 293T cells expressing the WT (NoV Rp) or mutant RdRp (Rp^GAA) and WT VPg (VPg) proteins. A plus symbol denotes addition of the corresponding plasmid. The samples were resolved on a 4-to-12% denaturing PAGE gel prior to Western blotting with anti-HA antibodies. The relevant masses from the standards are shown to the left of the Western blot image. The identities of the molecules are shown to the right of the Western blot image. VPg-RNA, VPg linked to RNA; VPg, free VPg. (B) Western blotting results for the total RNAs purified from HEK 293T cells that expressed either the WT or mutant RdRp and VPg proteins. The RNAs were electrophoresed in a 4-to-12% denaturing PAGE gel. A plus symbol in the rows labeled RNase A or RQ1 DNase denotes a 30-min treatment with a 10-μg final concentration or 5 units, respectively, of the appropriate enzyme prior to loading the sample for SDS-PAGE. VPg-RNA, VPg linked to RNA; VPg, free VPg. Lane 2 contains free VPg generated from cells transfected to produce only VPg and was loaded to serve as a molecular mass marker.

Fig. 7.

RNase L participates in the processing of the VPg-primed RNA products synthesized by the NoV RdRp in the NoV-5BR assay. (A) siRNA knockdown of RNase L decreased the levels of RNase L protein in HEK 293T cells. The cells were treated for 48 h with the concentrations of the siRNAs shown above the Western blot image. Twenty microliters of the cell lysate was then subjected to electrophoresis through a 4-to-12% denaturing PAGE gel. The blot was probed with antibodies to RNase L or β-actin, as shown to the right of the Western blot image. (B) Effects of RNase L knockdown on RIG-I signaling with exogenously supplied ligands or ligands produced by the NoV RdRp. The exogenously provided ligands were shR9 and S4dsRNA, produced by in vitro transcription using the T7 RNA polymerase and transfected into the cells using Lipofectamine. The NoV RdRp and VPg were transiently expressed from transfected plasmids. NS siRNA, a control, nonspecific siRNA (Invitrogen). The data are the means from three replicates and are representative of two independent experiments, and the standard errors are shown above the bars. Statistical analysis was performed by using the pairwise Student t test, and the P values are above the samples that were compared. (C) Knockdown of RNase L increased the abundance of the VPg-RNA complex made by the NoV RdRp. The Western blotting results are from immunoprecipitation reactions of cell lysates that expressed the NoV VPg without or with the NoV RdRp probed to detect the VPg protein. The lane VPg alone served as a molecular mass control for free VPg. (D) Confirmation of the effect of RNase L on the abundance of the VPg-RNA complex. Samples shown were used in the Western blot assay containing total RNAs purified from cells that were transfected to expressed the NoV RdRp and VPg after RNase L knockdown. The lane labeled VPg alone served as a molecular mass control for VPg that was not covalently linked to RNA. This species of VPg would not be present in the affinity purification of RNA.

Fig. 8.

Identification of a template for VPg-primed RNA synthesis in HEK 293T cells. (A) Alignment of the sequences of the cDNAs derived from the VPg-primed RNAs. The sequence from the cellular TGOLN2 mRNA is shown on the top line in the 5′-to-3′ direction, and the sequences of VPg-linked RNAs are complementary to and shown below the TGOLN2 mRNA in the 3′-to-5′ direction. Each sequence was from an independent cDNA clone, and the numbers in parentheses represent the number of independent sequences obtained. Dashes denote that the sequence was not present. (B) Northern blot of the VPg-RNA complexes immunoprecipitated from HEK 293T cells expressing the WT or mutant RdRp and VPg, performed with a radiolabeled oligodeoxyribonucleotide corresponding to the TGOLN2 mRNA sequence. Where RNase A or RQ1 DNase treatment was used, the immunoprecipitated material was incubated with the enzyme for 30 min prior to electrophoresis on a denaturing PAGE gel.

Fig. 9.

Effects of other NoV proteins on RdRp activity. (A) Western blotting of the NoV proteins showed their expression in HEK 293T cells. All of the proteins were HA tagged at their C termini and detected with the goat anti-HA antibody and an HRP-conjugated mouse anti-goat antibody. (B) Summary of the effects of coexpressed NoV GII.4 structural and nonstructural proteins on RIG-I signaling or NoV RdRp activity. Plasmid carrying NoV RdRp was transfected at 50 ng/well, and plasmid carrying each other protein was transfected at 10 ng/well. All data are presented as the means and standard errors of three replicates from three independent assays. The results are normalized to the vector-only control to better allow comparisons. Results statistically different from that of the vector-alone sample are shown with P values in parentheses. (C) Effects of the NoV proteins on the RdRps from NoV, MNV, HCV (1b5b), and BMV (BMV 2a). Plasmids encoding all RdRps were transfected into HEK 293T cells at 50 ng, and the plasmids that express the other NoV GII.4 proteins were transfected at 10 ng. Each bar represents the means and standard errors of three replicates from two independent assays. Statistical analysis was performed by using the pairwise Student t test, and the P values are indicated above the samples that were compared.