Environmentally-triggered contraction of the norovirus virion determines diarrheagenic potential

Abstract

Noroviruses are the leading cause of severe childhood diarrhea and foodborne disease worldwide. While they are a major cause of disease in all age groups, infections in the very young can be quite severe with annual estimates of 50,000-200,000 fatalities in children under 5 years old. In spite of the remarkable disease burden associated with norovirus infections in people, very little is known about the pathogenic mechanisms underlying norovirus diarrhea, principally because of the lack of tractable small animal models. We recently demonstrated that wild-type neonatal mice are susceptible to murine norovirus (MNV)-induced acute self-resolving diarrhea in a time course mirroring human norovirus disease. Using this robust pathogenesis model system, we demonstrate that virulence is regulated by the responsiveness of the viral capsid to environmental cues that trigger contraction of the VP1 protruding (P) domain onto the particle shell, thus enhancing receptor binding and infectivity. The capacity of a given MNV strain to undergo this contraction positively correlates with infection of cells expressing low abundance of the virus receptor CD300lf, supporting a model whereby virion contraction triggers infection of CD300lf^lo cell types that are responsible for diarrhea induction. These findings directly link environmentally-influenced biophysical features with norovirus disease severity.

Keywords: animal model; calicivirus; gastrointestinal infections; infectious diseases; norovirus.

Figure 1

Replication efficiency of viral chimeras. (A) Duplicate wells of BV2 cells were infected with MNV1, CR6, WU23 or the indicated chimeric virus at MOI 0.05. Virus titers were determined by TCID₅₀ assay on culture supernatants. Three independent experiments were performed. P values were determined using ANOVA with Dunnett’s multiple comparisons test by comparing each chimeric virus to its respective parental virus. (B) Groups of P3 BALB/c neonatal mice were inoculated with 10⁷ TCID₅₀ units of MNV1, CR6, WU23 or the indicated chimeric virus. At 2d, viral titers were determined in the distal small intestine by plaque assay. At least five mice from at least two independent litters per condition were analyzed. P values were determined using ANOVA with Sidak’s multiple comparisons test by comparing each chimeric virus to its respective parental virus. *P < 0.05, **P < 0.01, ****P < 0.0001.

Figure 2

The surface-exposed domain of the major capsid protein VP1 is the major MNV virulence factor in BALB/c mice. (A, B) Groups of P3 BALB/c neonatal mice were inoculated with 10⁷ TCID₅₀ units of mock inoculum, WU23, MNV1, CR6, or the indicated chimeric virus. (A) At 2 dpi, fecal consistency was determined by palpating their abdomens. (B) The proportion of mice scoring a 3 or above is presented as incidence of diarrhea. At least 5 mice from at least 2 independent litters per condition were analyzed. P values were determined for the MNV1/CR6 chimeras and the CR6/WU23 chimeras using Kruskal-Wallis test with Dunn’s multiple comparisons test by comparing each chimeric virus to its respective parental virus. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 3

VP1 is the major MNV virulence factor in C57BL/6J mice. Groups of P3 B6 pups were inoculated with 10⁷ TCID₅₀ units of mock inoculum, MNV1, WU23, CR6, or the indicated chimeric virus. (A) Fecal consistency was determined at 2 dpi by palpating their abdomens. (B) The proportion of mice scoring a 3 or above is presented as incidence of diarrhea. At least thirteen mice from a minimum of two different litters per condition were analyzed. Error bars denote standard errors of mean in all figures. P values were determined using Kruskal-Wallis test with Dunn’s multiple comparisons test comparing all groups to CR6. *P < 0.05, **P < 0.01.

Figure 4

Diarrheagenic MNV1 and WU23 can infect CD300lf^lo cells in vitro while attenuated CR6 cannot. (A, B) Duplicate wells of BV2 cells (A) or M12 cells (B) were infected with MNV1, CR6, or WU23 at MOI 0.05. Virus titers were determined by TCID₅₀ assay at the indicated timepoints on culture supernatants. P values were determined by ANOVA using Tukey’s multiple comparisons test. (C) Duplicate wells of M12 cells were infected with MNV1, CR6, or WU23 at MOI 10 or MOI 40 as indicated. Virus titers were determined by TCID₅₀ assay at the indicated timepoints. (D) CD300lf was overexpressed in M12 cells by lentivirus transduction. Duplicate wells of M12 cells or CD300lf-overexpressing M12 cells were infected with MNV1, CR6, or WU23, at MOI 5. Virus titers were determined by TCID₅₀ assay at 1 dpi. P values were determined using ANOVA with Sidak’s multiple comparisons test. At least three independent experiments were performed for all experiments. ****P < 0.0001.

Figure 5

Diarrheagenic MNV1 and WU23 are enhanced by a broader profile of environmental triggers than attenuated CR6. (A) BV2 cells were infected with 100 TCID₅₀ units per well of MNV1, WU23, or CR6 in neutral PBS, pH 7.6 or PBS supplemented with 10% FBS. Virus titers were determined by plaque assay at 2 dpi and normalized to viral titers following infection in media. P values were determined using ANOVA with Sidak’s multiple comparisons test. (B) BV2 cells were infected with 100 TCID₅₀ units per well of MNV1, WU23, or CR6 250 μM GCDCA resuspended in EtOH or vehicle control (VC). Virus titers were determined by plaque assay at 2 dpi and normalized to viral titers following infection in media. P values were determined using ANOVA with Sidak’s multiple comparisons test. (C) BV2 cells were infected with 100 TCID₅₀ units per well of MNV1 (left panel), WU23 (middle panel), or CR6 (right panel) in PBS at the indicated pH. Virus titers were determined by plaque assay at 2 dpi and normalized to viral titers following infection in media. P values were determined using ANOVA with Dunnett’s multiple comparisons test using pH 7.6 as the control group. At least three independent experiments were performed for all experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant.

Figure 6

While GCDCA causes comparable contraction of virulent and attenuated MNV strains, low pH causes variable conformational changes. (A, D, G) Sections of the image reconstructions of all three strains at pH 7.4 without bile or metals present. Note that all three have floating P domains. The linker between the shell and the P domain is more visible in MNV1 because it is a lower resolution structure where those details are not lost. (B, E, H) Magnified views of the MNV P domains at pH 7.4 in the presence of GCDCA (C, F, I) Representative images of a magnified view of the MNV P domains at pH 5.0. In the bottom CR6 image, arrow 1 indicates the weak density of the P domains and arrow 2 indicates the splaying of P1 domains at the base that is not observed for MNV1 or WU23.

Figure 7

CR6 differs from the other strains at low pH. (A, B) The rigid body fitting results of the structure of MNV1 into the unsharpened cryo-EM maps of CR6 at pH 5 in the absence (A) and presence (B) of 10mM GCDCA. Unlike the other strains, the angles between the P domains changes at low pH in the presence and absence of GCDCA. (C) To measure these angles, the center of masses of the P1 and P2 domains for both subunits were calculated (red spheres). These positions were then used to calculate the vectors between the subunits as summarized in panel d. (D) The vectors and angles between the P domains in the CR6 structure in the presence (red) and absence (green) of 10mM GCDCA are shown.

Figure 8

The hypervariable domain of VP1 regulates infection of CD300lf^lo cells and sensitivity to environmentally triggered infectivity enhancement. (A) Duplicate wells of M12 cells were infected with MNV1, CR6, WU23, or the indicated chimeric virus at MOI 0.05. Virus titers at 1 dpi were determined by TCID₅₀ assay on culture supernatants. All infections were repeated four times. P values were determined using ANOVA with Dunnett’s multiple comparisons test using CR6 as the control group. (B) BV2 cells were infected with 100 TCID₅₀ units per well of MNV1, WU23, CR6, CR6^VP1.MNV1, CR6^VP1.WU23, or CR6^P2VP1.WU23 in neutral PBS (pH 7.6) or PBS at pH 6.7. Virus titers were determined by plaque assay and normalized to viral titers following infection in media. P values were determined using ANOVA with Dunnett’s multiple comparisons test using CR6 as the control group. The experiment was repeated two times. *P < 0.05, **P < 0.01, ****P < 0.0001.