Discovery of a proteinaceous cellular receptor for a norovirus

Abstract

Noroviruses (NoVs) are a leading cause of gastroenteritis globally, yet the host factors required for NoV infection are poorly understood. We identified host molecules that are essential for murine NoV (MNoV)-induced cell death, including CD300lf as a proteinaceous receptor. We found that CD300lf is essential for MNoV binding and replication in cell lines and primary cells. Additionally, Cd300lf(-/-) mice are resistant to MNoV infection. Expression of CD300lf in human cells breaks the species barrier that would otherwise restrict MNoV replication. The crystal structure of the CD300lf ectodomain reveals a potential ligand-binding cleft composed of residues that are critical for MNoV infection. Therefore, the presence of a proteinaceous receptor is the primary determinant of MNoV species tropism, whereas other components of cellular machinery required for NoV replication are conserved between humans and mice.

Fig. 1. CRISPR screen identifies CD300lf as necessary for MNoV infection

(A) A heat map showing enrichment of genes in the two indicated conditions. Genes are color coded based upon their STARS score. (B) Wild-type BV2 cells or two independently-derived CD300lf-deficient clones (clone 1 and clone 2) transduced with an empty vector or a vector expressing CD300lf were challenged with an MOI of 0.05 of MNoV^CW3 (top) or MNoV^CR6 (bottom) infection and virus production was assessed by plaque forming units (PFU). Shown is mean ± SEM for data pooled from three independent experiments. L.O.D., limit of detection.

Fig. 2. CD300lf is an MNoV receptor

(A) Indicated BV2 cells were transfected with MNoV^CW3 RNA and harvested 12 hours post-transfection. Viral production was measured by plaque assay. ns = not statistically significant. Shown is mean ± SEM for data pooled from three independent experiments. (B) BV2 cells were incubated with either α-CD300lf or an isotype control prior to infection with MNoV^CW3 at an MOI 5.0. Cell viability was measured 24 hours post-infection. Shown is mean ± SEM for data pooled from three independent experiments. (C) MNoV^CW3 was incubated with either the recombinant, soluble ectodomain of CD300lf (sCD300lf) or a control protein prior to infection of BV2 cells. Cellular viability was assayed 24 hours post-infection. Shown is mean ± SEM for data pooled from three independent experiments. (D) MNoV^CW3 infection was inhibited by either cellular pretreatment of α-CD300lf or viral pretreatment with sCD300lf. Infection was measured by FACS for intracellular MNoV NS1/2 expression and inhibition relative to an isotype control or control protein, respectively. Shown is mean ± SEM for data pooled from three independent experiments. (E) MNoV^CW3 binding assay in complete media to indicated BV2 cell lines as assayed by bound MNoV genomes via qPCR. Representative binding assay when cells or virus were preincubated with α-CD300lf or sCD300lf, respectively. Analyzed by one-way ANOVA with Tukey’s multiple comparison test; three independent experiments performed in triplicate. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001. (F) MNoV^CW3 binding assay done in PBS plus 10% fetal bovine serum (FBS) or derivatives as indicated. Analyzed by one-way ANOVA with Tukey’s multiple comparison test; three independent experiments performed in triplicate. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Fig. 3. CD300lf is a physiological MNoV receptor

(A) MNoV^CW3 harvested from the spleens of BL6/J mice or MNoV^CW3 derived from BV2 cells was preincubated with sCD300lf prior to plaque assay. Inhibition is relative to a control protein. Shown is mean ± SEM for data pooled from three independent experiments. (B–C) Survival of Stat1^−/− mice from challenge with (G) 10³ PFU or (H) 10⁵ PFU of MNoV^CW3 preincubated with either sCD300lf or control protein. Analyzed by log-rank test; N=11 mice per cohort combined from three independent experiments. (D) Cd300lf^−/− and littermate control mice were challenged with 10⁶ PFU of MNoV^CR6 perorally. Fecal shedding of MNoV genomes was monitored for 21 days post-challenge. N=9–10 mice per cohort combined from two independent experiments. Analyzed by repeated measures ANOVA; shown is mean ± SEM. ***P<0.001.

Fig. 4. Structure guided mapping identifies the CC′-loop and CDR3 of CD300lf critical for MNoV infection

(A). HeLa cells transiently transfected with indicated constructs were infected with either MNoV^CW3 (left) or MNoV^CR6 (right) at an MOI of 0.05. Viral growth was measured by plaque assay at the indicated time points. Shown is mean ± SEM for data pooled from three independent experiments. (B) HeLa cells transiently transfected with indicated CD300 constructs or Tim1 were infected with MNoV^CW3 at an MOI of 5.0 and analyzed for expression of CD300/Tim1 (FLAG) and MNoV NS1/2. Shown is mean ± SEM for data pooled from three independent experiments. (C) Recombinant ectodomains of indicated CD300 molecules (10 μg/ml) were preincubated with MNoV^CW3 MOI 5 prior to infection of BV2 cells. Cellular viability was assayed 24 hours post-infection. Shown is mean ± SEM for data pooled from three independent experiments. (D) HeLa cells transiently transfected with indicated CD300 constructs were infected with MNoV^CW3 at an MOI of 5.0 and analyzed for expression of CD300 (FLAG) and MNoV NS1/2. CD300lfΔCT, CD300lf with truncation of cytoplasmic domain. Shown is mean ± SEM for data pooled from three independent experiments. (E) Ribbon diagram of murine CD300lf ectodomain with bound metal ion and HEPES. The β-sheets (cyan) are lettered as a canonical V-type Ig domain with the positions of the CDR equivalent loops indicated. The disulfide bonds are shown in yellow. (F) Mapping of mutational results onto the CD300lf surface. Displayed in cyan are murine CD300lf residues replaced by human equivalent residues in CDR1, C″, and DE-loop that had no effect on infection. The CC′-loop and CDR3 mutations that diminished viral infection are shown in magenta.