A Secreted Viral Nonstructural Protein Determines Intestinal Norovirus Pathogenesis

Abstract

Murine norovirus (MNoV) infects a low percentage of enteric tuft cells and can persist in these cells for months following acute infection. Both tuft-cell tropism and resistance to interferon-λ (IFN-λ)-mediated clearance during persistent infection requires the viral nonstructural protein 1/2 (NS1/2). We show that processing of NS1/2 yields NS1, an unconventionally secreted viral protein that is central for IFN-λ resistance. MNoV infection globally suppresses intestinal IFN-λ responses, which is attributable to secreted NS1. MNoV NS1 secretion is triggered by caspase-3 cleavage of NS1/2, and a secreted form of human NoV NS1 is also observed. NS1 secretion is essential for intestinal infection and resistance to IFN-λ in vivo. NS1 vaccination alone protects against MNoV challenge, despite the lack of induction of neutralizing anti-capsid antibodies previously shown to confer protection. Thus, despite infecting a low number of tuft cells, NS1 secretion allows MNoV to globally suppress IFN responses and promote persistence.

Keywords: IFN-λ; NS1; norovirus; secretion; vaccine.

Figure 1.. Global downregulation of ISG expression by CR6 infection in the intestine.

Wild-type mice were infected with 10⁶ plaque forming units (PFU) of CR6 perorally. At 3 dpi and 35 dpi, ileum from the infected mice and the littermate control mice were collected for mRNA sequencing. (A-D) Host gene expression profile was analyzed by GSEA. Net-enrichment-scores (NES) of negatively affected gene-sets by CR6-infection at 3 dpi (A) and 35 dpi (B). Enrichment plots of Interferon lambda response genes at 3 dpi (C) and 35 dpi (D). (E) A heat-map showing downregulation of ISGs by CR6 infection at 35 dpi. (F) Ifnlr1^f/f and Ifnlr1^f/f-Villincre mice were infected with 10⁶ PFU of CR6, and the expression of Ifit1, Oas1a and Ifi35 mRNA in ileum was analyzed at 35 dpi by qRT-PCR (n = 22-24 mice per group, combined from five independent experiments). Shown are means ± SEM. NS, not significant; *P < 0.05, **P < 0.01, determined by unpaired t-test. See also Figure S1.

Figure 2.. Caspase-3-mediated unconventional secretion of NS1 during MNoV infection.

(A) Schematic depicting NS1 protein maturation. (B) Immunoblots showing NS1 secretion during CR6 infection in BV2 cells. (C) Relative expression of NS1 and NS1/2 in the supernatant (sup) and in the lysate. Band intensity of immunoblots at 12 hpi was measured from five independent experiments. Shown are means ± SD. (D) HEK293T cells were transfected with Flag-NS1/2-CR6, and Flag-NS1 secretion was examined 48 hours after transfection. (E) NS1 secretion at 12 hpi with Caspase-cleavage site mutant viruses. (F) BV2 cells were treated with Q-VD-OPh (20 nM) at 4 hpi and NS1 secretion was examined at 12 hpi. (G) Casp3KO BV2 cells were generated by CRISPR, and were infected with CR6 to access NS1 secretion. (H) Intracellular flow cytometry plots. CR6-infected BV2 cells were co-stained with cleaved-Casp3 and NS6/7 at 16 hpi. (I) Brefeldin A (5 ng/ml) treatment to block the conventional secretion pathway. The cells were incubated with Brefeldin A from 4 hpi and NS1 secretion was analyzed at 12 hpi. (J) HEK293T cells were transfected with Flag-NS1/2^CR6 or Flag-NS1/2^GI.I and Flag-NS1^GI.1 cleavage and secretion was detected in the sup. Asterisk means non-specific bands. Representative data from at least three experiments are shown, except (I). See also Figure S2.

Figure 3.. Molecular characterization of secreted NS1

(A-B) Size exclusion chromatography with the culture supernatant from CR6-infected cells and rNS1 purified from E. coli. The samples were applied to Superdex 200 10/300 GL column for comparison. The corresponding fractions were collected in sequence as numbers indicated on the elution axis (B), then determined by either total protein staining or immunoblot with anti-NS1 and anti-capsid antibodies (A). (C) Molecular identification of secreted NS1. The sliced gel bands (Figure S3) were subjected to LC-MS/MS protein identification. The polypeptide fragments found in the MS/MS were mapped to each sample, respectively, as shown in the color used in (B). See also Figure S3.

Figure 4.. NS1 secretion is critical for mucosal infection but dispensable for systemic infection in vivo

(A) BV2 cells were inoculated with CR6 or CR6^D121/131G at MOI 0.1, and viral growth at 12 and 24 hpi was assessed by plaque assay. (B-C) Wild-type mice were infected with 10⁶ PFU of the indicated viruses perorally (PO) (B) or intraperitoneally (IP) (C), and analyzed at 3 dpi. MNoV genomes in stool, colon and spleen were quantified by qRT-PCR (n = 8–11 mice per group, combined from two independent experiments). NS, not significant; ***P < 0.001, determined by Mann-Whitney test (B and C).

Figure 5.. The defective viral growth of CR6^D121/131G is complemented in IFN-λ signaling deficient mice

(A-E) Wild-type and the knock out mice were infected with 10⁷ PFU of CR6^D121/131G perorally. (A) Complemented viral infection of CR6^D121/131G in Ifnlr1^−/− mice (n = 9-13, combined from three independent experiments). (B) Viral shedding of CR6^D121/131G at 3, 7 and 14 dpi in Ifnlr1^−/− mice (n = 5-6, combined from two independent experiments). (C) CR6^D121/131G stool shedding in Ifnlr1^f/f-Villincre mice. (D) Quantification of MNoV+ cells (NS1/2⁺NS6/7⁺) in IECs (CD45⁻EpCAM⁺) by flow cytometry. (E-F) n = 9-10, combined from two independent experiments. (E) MNoV-NS6/7 co-localizes with DCLK1, a tuft cell marker, in the colon of Ifnlr1^−/− mice infected with CR6^D121/131G at 14 dpi. Images are representative of one of at least three independent experiments. Dashed lines represent the epithelial barrier. White boxes in the overlaid image reflect the magnified inset images. Scale bars, 10 microns. *P < 0.05; **P < 0.01; ***P < 0.001, determined by Mann-Whitney test (A, C, and D) or two-way ANOVA (B).

Figure 6.. NS1-immunization prevents MNoV infection in vivo.

(A) Schematic outline of infection and immunization. (B-C) Concentration of α-NSI IgG (B) and α-capsid IgG (C) in the serum was measured by ELISA, and normalized by monoclonal antibodies CM79 and A6.2 respectively. (D-E) NS1-vaccination protected the mice from CR6-infection. MNoV genome from stool was quantified by qRT-PCR at 3 dpi (D) and 7 dpi (E). (F) VP1^P-dom immune-sera protect CR6-infection in BV2 cells. ID50 values for protection were determined by serial dilution of the immune-sera from 1:10 to 1:100,000. (G-H) Vaccination of NS1 or UV-CR6. MNoV genome from stool was quantified by qRT-PCR at 3 dpi (G) and 7 dpi (H). (I) Passive serum transfer protected the mice from CR6 infection. Serum from the immunized mice with OVA or NS1 was transferred to naïve mice 1 week prior to CR6 infection. MNoV genome from stool was quantified by qRT-PCR at 3 dpi. (B, C, F) n = 14-15 per group, combined from three independent experiments. (D-E) n = 19-20 per group, combined from four independent experiments. (G-H) n = 10 per group, combined from two independent experiments. (I) n = 9-10 per group, combined from two independent experiments. Shown are means ± SEM. NS, not significant; *P < 0.05; **P < 0.01; ***P < 0.001, determined by Mann-Whitney test (B-C infected group, I) or Kruskal-Wallis test (B-C immunized group, D-H). See also Figure S4 and Figure S5.