Norovirus-Mediated Modification of the Translational Landscape via Virus and Host-Induced Cleavage of Translation Initiation Factors

Abstract

Noroviruses produce viral RNAs lacking a 5' cap structure and instead use a virus-encoded viral protein genome-linked (VPg) protein covalently linked to viral RNA to interact with translation initiation factors and drive viral protein synthesis. Norovirus infection results in the induction of the innate response leading to interferon stimulated gene (ISG) transcription. However, the translation of the induced ISG mRNAs is suppressed. A SILAC-based mass spectrometry approach was employed to analyze changes to protein abundance in both whole cell and m7GTP-enriched samples to demonstrate that diminished host mRNA translation correlates with changes to the composition of the eukaryotic initiation factor complex. The suppression of host ISG translation correlates with the activity of the viral protease (NS6) and the activation of cellular caspases leading to the establishment of an apoptotic environment. These results indicate that noroviruses exploit the differences between viral VPg-dependent and cellular cap-dependent translation in order to diminish the host response to infection.

Fig. 1.

A defect in ISG protein synthesis, but not mRNA induction is observed during norovirus infection. (A) Genome schematic of the norovirus genome with the four previously described murine norovirus open reading frames are shown (11). The NS1/2–7 nomenclature for the mature peptides generated from ORF1 (described in (8)) is used throughout. (B) Western blot for the norovirus NS7 nonstructural protein and the ISGs STAT1, ISG15, and viperin. (C) qRT-PCR for norovirus genomic RNA and (D–F) for the ISGs STAT1, ISG15, and viperin. The samples for panels (B–F) were taken from a high multiplicity of infection (MOI: 10 TCID₅₀/cell) timecourse performed in RAW 264.7. The correct position for viperin is indicated with “<” and is immediately beneath the visible nonspecific band. For qRT-PCR n ≥ 3, and error bars represent standard deviation.

Fig. 2.

MNV infected cells respond to interferon at the transcriptional but not posttranscriptional level. Samples were harvested from RAW 264.7 cells, infected at high MOI (10), 18 h post infection following treatment with interferon or mock culture supernatant. qRT-PCR of norovirus and ISG mRNA levels following interferon treatment (A–D) and representative Western blots are shown in (E), with densitometry analysis of protein levels in (F–I) error bars show standard deviation, n = 3. Data are presented relative to a mock, untreated control. Statistical analysis was performed by one-way ANOVA, with the exception of figure (F), which was performed by t test. For qRT-PCR, statistical analysis was performed on the untransformed ΔΔCt data. (* = <0.05, ** = <0.01, *** = <0.001, **** = <0.0001).

Fig. 3.

Norovirus infection alters the translational profile of the host cell. (A) RAW 264.7 or (B) BV-2 cells infected at an MOI of 10 TCID₅₀/cell with MNV were pulse labeled for 1h at the indicated time points and imaged on a phosphorimager. Polysome profiling of infected (C) RAW 264.7 or (D) BV-2 cells was performed. (E) Puromycylation analysis of infected BV-2 cells shows some minor enrichment of sites of active translation colocalizing with viral replication complexes visualized using anti-dsRNA.

Fig. 4.

Quantitative proteomic analysis of translation initiation during MNV infection. SILAC-based quantitative proteomics was employed to investigate changes to eIF composition during high MOI (10) MNV infection of BV-2 cells. The experimental layout is illustrated in (A) with samples taken at mock (0 h) early (4 h) or late (9 h) post infection either lysed or subject to m7GTP-Sepharose purification. A Coomassie gel (B) and representative Western blots (C) confirm initiation factor enrichment. Venn diagrams illustrating experimental proteome coverage in (D) whole-cell lysate, or E) m7GTP-Sepharose pulldowns.

Fig. 5.

Norovirus infection alters the abundance and eIF4F association of cellular translation initiation factors. Mass spectrometry data for individual eIF components identified in the whole cell (WCL) or m7GTP-Sepharose (m7GTP) experiments are shown, including (A) eIF4E, (B) eIF4A, (C) eIF4B, and (D) eIF4G. The whole eIF3 complex was successfully identified by mass spectrometry and its relative abundance in (E) WCL or (F) m7GTP samples is shown. Significance was tested by one-way ANOVA comparing changes to a control protein with unaltered abundance (eIF4E). Changes in eIF4E levels were determined by comparing its 4 h and 9 h levels. Error bars represent standard deviation, * = <0.05, ** = <0.01, *** = <0.001, **** = <0.0001. Where a protein was identified in only a single mass spectrometry replicate, precluding statistical analysis, this is indicated with “N.D.,” otherwise proteins were identified in at least 2/3 replicates.

Fig. 6.

Cleavage of PABP by the norovirus protease NS6 contributes to reduced cellular translation. (A) Illustration of the domain structure of PABP. (B) Western blot analysis of selected eIF proteins in 293T cells transfected with the MNV protease NS6. (C) Western blot analysis of PABP cleavage over a MOI 10 infection timecourse in BV-2 cells. (D) Analysis of global translation in 293T cells transfected with NS6 assessed by ³⁵S-methionine pulse-labeling and quantification on a phosphorimager. Error bars represent standard deviation from three biological replicates. Statistical analysis was performed by one-way ANOVA. (* = <0.05, ** = <0.01, *** = <0.001, **** = <0.0001).

Fig. 7.

Modulation of PABP cleavage during MNV infection is inhibitory for viral replication. (A) Western blot analysis of BV-2 cells infected at high MOI (10) with MNV and transfected with wild-type or noncleavable (Q440A) PABP. Viral titers obtained following low MOI (0.01) infection of (B) BV-2 or (C) RAW 264.7 cells transfected with wild-type or a noncleavable form (Q440A) of PABP. (D) Viral titers obtained following low MOI (0.01) infection of cells heterozygous for a truncated form of PABP. Error bars represent standard deviation from three biological replicates. Statistical analysis was performed by one-way ANOVA. (* = <0.05, ** = <0.01, *** = <0.001, **** = <0.0001).

Fig. 8.

Induction of apoptosis and caspase cleavage of eIF4F components also contributes to altered translation and alters eIF4F composition. (A) A diagram illustrating the structure of eIF4GI and II as well as their caspase cleavage sites. (B) Quantification of peptides mapping to the N-FAG, M-FAG, or C-FAG domains of eIF4GI binding m7GTP-Sepharose beads at 9 h post infection. (C) Western blotting against eIF4G and markers of apoptosis for an infection timecourse from BV-2 or (D) RAW 264.7 cells. (E) Western blotting of eIF4GI or II, and viperin in the presence of varying amounts of the caspase inhibitor z-vad-fmk. The specific viperin band is highlighted with “<” (F) ³⁵S-Methionine pulse labeling of BV-2 cells mock- or infected with MNV at 9 h post infection in the presence or absence of the z-vad-fmk inhibitor. (G). Quantification by phosphorimaging of three biological repeats of experiment shown in (F). (H). Viral titers obtained following low MOI (0.01) infection of BV-2 cells treated with z-vad-fmk (20 μm), necrostatin-1 (40 μm) singly or in combination. n ≥ 3, error bars represent standard deviation. Statistical analysis was performed by one-way ANOVA. (* = <0.05, ** = <0.01, *** = <0.001, **** = <0.0001).

Fig. 9.

Model for modifications to eIF4F during norovirus infection. (A) In healthy cells, the 5′ cap of mRNA is bound by eIF4E. This is bound by the scaffolding protein eIF4G that allows binding to other initiation factors (eIF4A, eIF3, PABP) as well as recruitment to the ribosome via the eIF3 complex. The norovirus protease NS6 alone can inhibit translation through cleavage of PABP, though larger-scale alterations to the initiation factor complex are not observed. In infected cells, the induction of apoptosis can result in further modification of the eIF complex with caspase cleavage of (C) eIF4GI separating the eIF4E and PABP-binding regions and abolishing the circularization of translating mRNAs. (D) Cleavage of eIF4GII by caspases is more extensive and in addition to the effects observed with eIF4GI cleavage, separates the eIF4E and eIF3-binding domains of eIF4GII, preventing recruitment of mRNA to the ribosome.