CD300lf is the primary physiologic receptor of murine norovirus but not human norovirus.

Murine norovirus (MNoV) is an important model of human norovirus (HNoV) and mucosal virus infection more broadly. Viral receptor utilization is a major determinant of cell tropism, host range, and pathogenesis. The bona fide receptor for HNoV is unknown. Recently, we identified CD300lf as a proteinaceous receptor for MNoV. Interestingly, its paralogue CD300ld was also sufficient for MNoV infection in vitro. Here we explored whether CD300lf is the sole physiologic receptor in vivo and whether HNoV can use a CD300 ortholog as an entry receptor. We report that both CD300ld and CD300lf are sufficient for infection by diverse MNoV strains in vitro. We further demonstrate that CD300lf is essential for both oral and parenteral MNoV infection and to elicit anti-MNoV humoral responses in vivo. In mice deficient in STAT1 signaling, CD300lf is required for MNoV-induced lethality. Finally, we demonstrate that human CD300lf (huCD300lf) is not essential for HNoV infection, nor does huCD300lf inhibit binding of HNoV virus-like particles to glycans. Thus, we report huCD300lf is not a receptor for HNoV.


Introduction
Human norovirus (HNoV) is the leading cause of infectious gastroenteritis globally, yet our understanding of how HNoV enters cells is limited [1][2][3][4]. Viral entry is the first and often ratelimiting step of the viral life cycle and is a major determinant of cell tropism, species range, host genetic susceptibility, and pathogenesis [1,5]. Entry includes virion attachment to the host cell membrane, receptor engagement, internalization, and genome release into the cell cytoplasm [1,5]. Both histo-blood group antigens (HBGAs) and bile salts bind HNoV and promote infection [6][7][8]. The identity of a proteinaceous cellular receptor(s) for HNoV remains unknown [1,8,9].
Murine norovirus (MNoV) represents a surrogate animal model for studies of HNoV infection and pathogenesis [1,10,11]. In contrast to HNoV, diverse infectious molecular clones exist for MNoV that can be readily propagated both in vitro and in mice. MNoV and HNoV share similar genome and capsid architecture, fecal-oral transmission routes, interactions with bile salts, and the ability to cause both acute and persistent infection [1,7,8,[10][11][12][13][14]. Recently, we identified CD300 family members CD300ld and CD300lf as functional receptors for MNoV [15,16]. CD300lf is necessary for infection of MNoV-susceptible RAW 264.7 and BV2 cell lines while both CD300ld and CD300lf are sufficient to confer susceptibility to cell lines from different species when ectopically expressed [15,16]. In addition, CD300lf-deficient mice are resistant to fecal-oral transmission of MNoV strain CR6 (MNoV CR6 ) [15]. Whether CD300lf is essential for genetically diverse MNoV strains, parenteral infection routes, or in the setting of immunodeficiency is unclear. Finally, it is unknown whether human CD300lf functions as a receptor for HNoV.
Noroviruses are non-enveloped, positive-sense single-stranded RNA viruses [1]. The viral capsid exhibits T = 3 icosahedral symmetry and is comprised of 90 dimers of the major structural protein VP1 [17]. VP1 contains a shell (S) domain and a protruding (P) domain that contains a proximal P1 and distal P2 subdomain [18]. P2 is the target for several neutralizing monoclonal antibodies for both HNoV and MNoV and contains the receptor binding site for MNoV [12,13]. Co-crystal structures and mutagenesis studies have identified that the CC' and CDR3 loops of CD300lf directly bind to a cleft between the AB' and DE' loops of the MNoV VP1 P2 subdomain [13,15,19]. Each CD300lf binds one P2 monomer, albeit at relatively low affinity, suggesting that norovirus-receptor interactions are largely driven by avidity [13]. Interestingly, binding of MNoV to CD300lf is promoted by both divalent cations and bile salts [12,13,15].
Diverse MNoV strains have been described with distinct pathogenic properties [20][21][22]. Specifically, MNoV CW3 causes acute systemic infection that is cleared by the adaptive immune system, while other strains including MNoV CR6 cause chronic enteric infection that can persist for months, if not the life of the animal [20,23]. Interestingly, a single amino acid in the P2 domain of MNoV VP1 is sufficient to confer lethality in mice deficient in type I interferon signaling, suggesting a role for virus-receptor interactions in determining pathogenesis [20,24].
There are eight CD300 family members in mice and seven in humans, each containing a conserved ectodomain, a single transmembrane domain, and a more variable cytoplasmic signaling domain [25,26]. The ectodomains bind diverse phospholipids found on dead and dying cells, resulting in activating or inhibitory signals depending upon the cellular and molecular context [26][27][28][29][30]. Both CD300ld and CD300lf are expressed in diverse myeloid cells, while CD300lf is additionally expressed in lymphoid cells and rare intestinal epithelial cells called tuft cells, which are the primary target cell of MNoV CR6 [25,31].
Several important questions remain regarding norovirus entry: is CD300lf the only receptor in vivo? Do other CD300 family members such as CD300ld contribute to infection? Does HNoV use CD300lf as a receptor? Here we show that CD300lf is essential for fecal-oral transmission and pathogenesis of diverse MNoV strains in both immunocompetent and immunodeficient mice, suggesting CD300lf is the primary physiologic receptor for MNoV. We also identify CD300lf-independent viral infection in the setting of extra-intestinal challenge in immunodeficient mice, suggesting a potential role for CD300ld or other receptors in this nonphysiological infection route. Finally, we demonstrate that CD300lf and related CD300 family members do not function as HNoV receptors.

Results
HNoV induces vomiting and diarrhea within approximately 24 hours after infection [32][33][34]. In contrast, MNoV replicates to high titers but is largely avirulent in immunocompetent mice [10,35,36]. We previously demonstrated that Cd300lf -/mice do not shed detectable MNoV CR6 between 3 and 21 days after per oral (PO) challenge; however, the role of CD300lf and potential role of CD300ld during earlier time points and in systemic tissues is unknown [15]. MNoV CW3 is an infectious molecular clone derived from the MNV-1 plaque isolate CW3 [14,20]. MNoV CW3 can infect myeloid and lymphoid cells in the intestine and secondary lymphoid organs and cause lethal infection in mice deficient in type I interferon signaling [14,20,37,38]. To test whether CD300lf was essential for MNoV CW3 infection ex vivo, we challenged bone marrow-derived macrophages (BMDMs) from Cd300lf +/or Cd300lf -/mice with MNoV CW3 . Cd300lf -/-BMDMs did not produce infectious virus as measured by plaque assay (Fig 1A) nor did they express the MNoV non-structural protein NS1/2 as detected by flow cytometry (Fig 1B and 1C). CD300lf is thus essential for MNoV CW3 infection of murine BMDMs. To test the role of CD300lf in vivo, we challenged Cd300lf +/and Cd300lf -/littermate mice with 10 6 plaque forming units (PFU) of MNoV CW3 perorally (PO) and measured infectious virus by plaque assay (Fig 1D). We also quantified viral genomes by qPCR (Fig 1E). At 24 hours post-infection (hpi), infectious virions were undetectable in Cd300lf -/mice in the mesenteric lymph node (MLN), ileum, and colon, in contrast to Cd300lf +/littermates. At this dose and time point, MNoV CW3 was not detected in the spleen of control or knockout mice. Viral genomes were similarly undetectable in Cd300lf-deficient animals ( Fig 1E).
Different MNoV strains have different in vivo tissue and cellular tropism, supporting the possibility that they may utilize distinct cellular attachment and entry mechanisms [31,38,39]. Next, we asked whether CD300lf was required for infection by genetically diverse MNoV strains. MNoV strains are derived from a single genogroup (Genogroup V) and alignments of MNoV VP1 previously revealed at least five distinct clusters, from which we selected representative strains MNoV CW3 , MNoV CR6 , MNoV WU23 , MNoV CR3 , and MNoV CR7 (Table 1) [21,[40][41][42]. An alignment of strains highlights conserved regions of the capsid protein, including the CD300lf binding sites and bile-acid interacting sites described previously (S1 Fig) [13]. Among these strains, there is 84% amino acid identity in the P domain and 98% amino acid identity in the more conserved S domain. The 11 VP1 residues shown to interact with the secondary bile salts glycochenodeoxycholic acid (GCDCA) or lithocholic acid (LCA) are perfectly conserved [13]. However, there is variation in 4 of the 17 VP1 residues previously shown to interact with CD300lf [13].
These data suggest that diverse strains may utilize receptors other than CD300lf (S1 Fig) [13]. To test the hypothesis that diverse MNoV strains differentially utilize CD300lf, we first  tested whether CD300lf was necessary and sufficient for infection by these strains in vitro. We challenged CD300lf-deficient BV2 microglial cells, which are pan-susceptible to all known MNoV strains, with six diverse MNoV strains. All strains induced cytopathic effects in wildtype BV2 cells but not CD300lf-deficient cells, demonstrating CD300lf is essential for virus induced-lethality in BV2 cells (Fig 2A). Next, we asked whether CD300ld and CD300lf are sufficient to confer susceptibility to human HeLa cells, which are not normally susceptible to MNoV as they lack expression of murine CD300 proteins. Consistent with our prior findings, both CD300ld and CD300lf are sufficient for infection by MNoV CW3 as measured by expression of the MNoV non-structural protein NS6/7 by flow cytometry [15]. Similarly, both CD300ld and CD300lf are sufficient for MNoV WU23 , MNoV CR3 , MNoV CR7 , MNoV MNV3 , and MNoV S99 infection ( Fig 2B). All tested MNoV strains utilize CD300ld and CD300lf at similar efficiencies when overexpressed in HeLa cells.
To determine whether CD300lf was essential for infection in vivo, we challenged Cd300lf +/and Cd300lf -/littermates with 10 6 PFU PO of either MNoV CW3 , MNoV CR6 , MNoV WU23 , MNoV CR3 , or MNoV CR7 and measured viral genomes seven days post-infection (dpi). These viruses represent each of the five distinct MNoV VP1 phylogenetic clusters described previously [40]. MNoV CW3 causes acute systemic infection that is cleared by 10-14 dpi, while MNoV CR6 , MNoV CR3 , and MNoV CR7 cause persistent infection [21]. The in vivo characteristics of MNoV WU23 are uncharacterized [21]. Consistent with prior studies, we observed differential tissue tropism associated with different MNoV strains [21]. In CD300lf +/control mice, viral genomes for all tested MNoV strains were shed in the feces of some mice with the exception of MNoV CW3 (Fig 2C). All tested MNoV strains were detected in the MLN ( Fig 2D). All strains were detectable albeit at varying levels in the ileum and colon. Interestingly, only MNoV CW3 and MNoV WU23 were detectable in the spleen. MNoV CR6 , MNoV CR3 , and MNoV CR7 , which are known to establish chronic infection in immunocompetent mice did not have detectable viral genomes in the spleen. MNoV genomes were not detected in any tissue (MLN, spleen, ileum, colon, or feces) of any Cd300lf-deficient mice challenged with any of the five MNoV strains (Fig 2C-2G). To determine whether there was MNoV infection below the limit of detection, we collected feces from Cd300lf +/or Cd300lf -/mice challenged with MNoV CR6 seven days prior, and then gavaged the feces into Stat1 -/recipient mice which lack innate immunity and are exquisitely sensitive to MNoV infection [14,15]. Seven days after fecal gavage we detected MNoV CR6 in Stat1 -/mice challenged with feces from Cd300lf +/mice but not from Cd300lf -/mice ( Fig 2H). These results suggest CD300lf is critical for infection by diverse MNoV strains in immunocompetent mice in vivo.
To further determine whether CD300ld or an alternative receptor was sufficient for MNoV infection below our limit of detection, we assessed the humoral immune response against MNoV CW3 at 14 days dpi, a time point after viral clearance by the adaptive immune system. A 1:10 dilution of sera from Cd300lf +/mice that cleared MNoV CW3 infection neutralized MNoV CW3 in vitro as measured by BV2 cell viability ( Fig 3A). The IC 50 from Cd300lf +/mice was approximately 1:80 in contrast to sera from Cd300lf -/mice which did not reach an IC 50 ( Fig 3B). To determine whether MNoV CW3 elicited a non-neutralizing humoral response in Cd300lf -/mice, we measured anti-MNoV IgG and IgM in the sera. Cd300lf +/mice had a significantly higher anti-MNoV IgG and IgM response consistent with resistance of Cd300lf -/mice to MNoV CW3 infection (Fig 3C and 3D).

PLOS PATHOGENS
CD300lf is the primary receptor of murine norovirus Cd300lf +/mice, but this was not statistically significant in the MLN, ileum, colon, or feces (Fig 4C-4F). Relative to Stat1 +/+ mice, viral genome copies were elevated and extraintestinal spread to the spleen was observed consistent with prior findings (Fig 2C-2G). Importantly, Cd300lf -/-Stat1 -/mice were resistant to MNoV CR6 in all tissues examined. Two

PLOS PATHOGENS
CD300lf is the primary receptor of murine norovirus Cd300lf -/-Stat1 -/mice had detectable viral genomes in the MLN but not in other tissues ( Fig  4C). A third mouse had detectable low-level viral RNA in the colon but not in any other tissue, possibly reflective of input virus or a false positive (Fig 4F).
The natural defenses of the gastrointestinal mucosa represent a bottleneck to viral infection. To test whether CD300lf was essential for parenteral infection routes, we bypassed the gastrointestinal tract by administering 10 7 PFU MNoV CW3 intraperitoneally (IP) to Cd300lf -/-Stat1 -/mice. All Cd300lf +/+ Stat1 -/and Cd300lf +/-Stat1 -/mice died by five or six dpi, respectively ( Fig  5A). Two of nine Cd300lf -/-Stat1 -/mice died at 10 and 14 dpi with the remaining mice surviving at least 21 days. The two Cd300lf -/-Stat1 -/mouse deaths were observed in independent experiments. Virological data was not available from these animals; therefore, whether MNoV CW3 contributed to the death of these two Cd300lf -/-Stat1 -/mice is unknown.
Next, given the structural and genetic similarity between MNoV and HNoV, we tested whether human CD300lf (huCD300lf) is a receptor for HNoV. First, we assessed the ability of huCD300lf to prevent binding of recombinant HNoV virus-like particles (VLPs) to HBGAs in pig gastric mucin. To increase the avidity of the potential CD300lf and VLP interaction, we generated Fc-fusion proteins with the either huCD300lf or human CD300ld (huCD300ld) ectodomains. We recently demonstrated that an Fc-fusion protein of mouse CD300lf has substantially increased binding and neutralizing ability on MNoV [13]. We incubated HNoV VLPs with 10 μg/ml of either Fc-huCD300lf, Fc-huCD300ld, or a buffer-only control and measured HBGA-bound VLPs by ELISAs. Neither Fc-huCD300lf or Fc-huCD300ld inhibited binding of GI.1 Norwalk, GI.3 Desert Shield Virus, GII.4 1997, or GII.4 2012 VLPs to HBGAs (Fig 6A) [43,44]. Next, we asked whether potential HNoV interactions with huCD300lf were necessary for infection of human intestinal enteroids (HIEs). Differentiated monolayers of HIEs were incubated with polyclonal anti-huCD300lf or an IgG1 control and then challenged with HNoV GII.4 from stool filtrate. We previously demonstrated that a similar polyclonal anti-muCD300lf potently inhibited MNoV infection [15]. Anti-huCD300lf had no effect on HNoV genome replication at 24 hpi relative to the isotype control ( Fig 6B). Finally, we asked whether Fc-huCD300lf could neutralize HNoV GII.4. Virus was pre-incubated with up to 50 μg/ml of Fc-huCD300lf and then used to infect HIEs. Fc-huCD300lf did not affect viral genome replication relative to a control protein (Fig 6C). Together these data indicate that huCD300lf is not a receptor for HNoV.

Discussion
Here we demonstrate that CD300lf, and not CD300ld, is the primary physiologic receptor for diverse MNoV strains in vivo. We demonstrated that CD300lf is essential for detectable viral

PLOS PATHOGENS
CD300lf is the primary receptor of murine norovirus infection at multiple time points with diverse MNoV strains as measured by plaque assay, qPCR, and serologic response. Interestingly, both CD300ld and CD300lf are sufficient to mediate MNoV entry when ectopically expressed in vitro yet CD300ld is not sufficient for oral or parenteral transmission in immunocompetent mice. Several possibilities may explain this discordance. First, CD300ld receptor utilization has only been described when overexpressed at supraphysiologic levels [15,16]. Similarly the soluble recombinant ectodomain of CD300lf can neutralize MNoV infection in contrast to that of CD300ld [15]. This raises the possibility that CD300ld utilization by MNoV is inefficient relative to CD300lf and that our overexpression assays are not sufficiently sensitive to quantify differences in receptor utilization efficiency. A similar phenomenon has been described with HIV-1, which can engage co-receptors other than CCR5 and CXCR4 when overexpressed in cell lines, but not under physiologically relevant conditions [45,46]. Second, CD300ld and CD300lf have overlapping yet distinct cell type expression, raising the possibility that CD300ld-expressing cells are either not permissive to MNoV because of a post-entry restriction or because CD300ld-expressing cells are not anatomically accessible to MNoV. Consistent with this, intestinal tuft cells, which are the major target of MNoV CR6 in the intestines, express CD300lf but not CD300ld [31,47].
Although we demonstrated that Cd300lf-deficient mice are resistant to diverse MNoV strains via both oral and parenteral routes, we observed detectable viral genomes in Cd300lf -/-Stat1 -/mice inoculated with high-dose MNoV CR6 IP, consistent with CD300lf-independent viral replication. Sanger sequencing of virus from tissues from multiple mice revealed a single amino acid mutation (F514I) in the P1 domain of MNoV VP1 of Stat1 -/mice with and without a functional copy of CD300lf, suggesting this variant emerged independent of CD300lf interactions. This raises the question as to why this viral variant emerged and whether it has physiologic relevance, to be explored in future studies.
The physiologic function of CD300lf is as a cell death sensor and immunoregulatory protein [25]. Interestingly, the CD300 locus is under positive selection and there are a number of CD300lf polymorphisms in laboratory and wild mouse strains that may affect CD300lf expression or conformation and thus may regulate susceptibility to MNoV [48]. This raises the intriguing hypothesis of a molecular arms race between MNoV and mice over evolutionary

PLOS PATHOGENS
CD300lf is the primary receptor of murine norovirus time which may have contributed to the relatively non-pathogenic consequences of MNoV in immunocompetent mice.
Finally, these data have several important implications for our understanding of HNoV, the entry mechanisms of which remain to be defined. Given the structural similarity between MNoV and HNoV as well as mouse and human CD300lf, we hypothesized that human CD300lf might be a functional receptor for HNoV. However, via multiple orthogonal approaches we demonstrated that human CD300lf is not a functional receptor for several GI and GII HNoVs. This further raises the question as to the identity of the HNoV receptor and whether or not it is structurally related to CD300lf. This remains an important area of future investigation that would significantly enhance our understanding of HNoV pathogenesis and facilitate novel prophylactic and therapeutic approaches.

PLOS PATHOGENS
CD300lf is the primary receptor of murine norovirus

Mouse strains
All mouse strains used in this experiment were from a C57BL/6J background (Jackson Laboratories, Bar Harbor, ME) and bred in-house. Generation of B6.CD300lf em1Cbwi /J (Cd300lf -/-) mice (Jackson Laboratories) and B6.129S(Cg)-Stat1 tm1Div /J (Stat1 -/-) (gift of H.W. Virgin) were previously described [15,49]. These mice were housed in a MNoV-free facility at Yale University School of Medicine. All experiments used littermate controls, were gender-balanced, and done at least two independent times. Mice were used for infections between 6-10 weeks of age. Genotyping of mice was done by real time PCR as described previously [15].

Ethics statement
The care and use of the animals were approved by and in agreement with the Yale Animal Resource Center and Institutional Animal Care and Use Committee (#2018-20198) according to standards in the Animal Welfare Act.

PLOS PATHOGENS
CD300lf is the primary receptor of murine norovirus

In vitro MNoV infections
BV2-WT and BV2ΔCD300lf cells (gift of H.W. Virgin) were seeded at 20,000 cells per well in 96-well plates and infected with diverse MNoV strains at a MOI of 5. After 48 hours, 25μL of CellTiter-Glo (Promega, Madison, WI) was added to each well and luminescence was detected on a Synergy luminometer (BioTek). Experiments were performed in duplicate in at least three independent experiments. BMDMs were infected for 24 hours with MNoV CW3 at a MOI 5 for flow cytometry experiments and a MOI 0.05 for plaque assays. Human enteroids were inoculated with HNoV GII.4 from stool filtrate as described previously [8]. Tranfected 293T cells were challenged with HNoV GII.6 from stool filtrate as described below. The HNoV GII.6 stool filtrate was determined to be infectious using BJAB cells as described previously [9].

In vitro HNoV infections in HIEs
Five-day differentiated jejunal HIE monolayers in 96-well plates were inoculated with GII/Hu/ US/2012/GII.4 Sydney [P31]/TCH12-580 (2.5x10 5 genome copies/well) from stool filtrate as described previously [8]. This virus has no accession number as it is from a stool filtrate. Experiments were conducted in triplicate in three independent experiments. Serial dilutions of polyclonal antibody against the huCD300LF ectodomain (R&D Systems; AF2774) or recombinant human CD300LF/LMIR3 Fc chimera protein (R&D Systems; 2774-LM-050) samples were carried out in CMGF(-) medium containing 500 μM glycochenodeoxycholic acid (GCDCA). An isotype control (Recombinant Human IgG1 Fc, CF; R&D Systems; 110-HG-100) and recombinant rotavirus NSP4 protein were used as controls respectively. GII.4_Sydney_2012 (2.5x10 5 genome copies) were mixed with an equal volume of media or dilutions of each serum or protein sample at 37˚C for 1 hr, and inoculated onto jejunal HIE monolayers for another 1 hr at 37˚C in 5% CO 2 . After incubation, monolayers were washed twice with CMGF(-) media to remove unbound virus and cultured in differentiation media with 500 μM GCDCA for the indicated time points. RNA was extracted from each well using the KingFisher Flex Purification system and MAgMAX-96 Viral RNA Isolation kit. RNA extracted at 1 hpi, was used as a baseline to determine the amount of input virus that remained associated with cells after washing the infected cultures to remove the unbound virus. Replication of virus was determined by RNA levels quantified from samples extracted at 24 hpi.

Mouse infections
Mice were perorally (PO) inoculated with 25μL of 10 6 PFU MNoV diluted in D10 (DMEM with 10% FBS). For intraperitoneal (IP) challenge, MNoV was diluted to either 10 7 PFU per 200μL in phosphate buffered saline (PBS) and injected into the left lower quadrant of the peritoneal cavity with an insulin syringe. Fecal transmission assay was performed by infecting Cd300lf -/or Cd300lf +/mice with 25μL PO of 10 6 PFU MNoV CR6 . At seven dpi, a freshly isolated fecal pellet was homogenized in 100μL of PBS and 25μL of the fecal slurry was administered PO to Stat1 -/mice. Seven days after fecal transfer, fecal samples from the Stat1 -/recipient mice were tested for MNoV genomes via qPCR as described below.

Virus quantification by plaque assay
BV2 cells were seeded in 6-well plates at 2 x10 6 cells/well. Tissue samples were weighed and homogenized with 1.0 mm silica beads (BioSpec, Bartlesville, OK) in 1 mL of D10 with 1% Pen/ Strep, and 1% HEPES [38]. Tissue homogenates were serially diluted ten-fold. Media was aspirated off the BV2 cells and samples were inoculated to each well and gently rocked for 1 hour. Inoculum was removed and 2 mL of overlay media (MEM containing 1% methylcellulose, 10%

PLOS PATHOGENS
CD300lf is the primary receptor of murine norovirus FBS, 1% GlutaMAX (Gibco), 1% HEPES, and 1% pen/strep) was added to each well. Inoculated plates were incubated for 48 hours at 37˚C and 5% CO 2 prior to plaque visualization with crystal violet (0.2% crystal violet in 20% ethanol) as described previously [37]. Plaque assay from cell culture was performed by freeze/thawing infected samples at 0, 12, and 24 hpi followed by serial dilutions as described above [15,37].

Flow cytometry
BMDMs were harvested after mock-inoculation or infection with MNoV. Cells were pelleted by centrifugation at 200 x g for 5 min and suspended in 250μL of Cytofix/Cytoperm (BD Biosciences) for 20 min at RT. Cells were washed twice with perm/wash buffer (PWB), suspended in staining buffer containing rabbit anti-NS1/2 antibody (1:1000; kind gift of Vernon Ward). The cells were incubated with antibody for 30 min at RT, pelleted and washed twice. The cells were suspended in staining buffer containing donkey anti-rabbit alexa fluor 647 (1:500, Life Technologies #A31573) and incubated for 30 min in the dark. The cells were pelleted and washed twice with PWB buffer. Cells were suspended in FACS buffer. Flow cytometry was performed on a MACSQuant Analyzer 10 (Miltenyi Biotec, Somerville, MA) and analyzed using FlowJo v10 (FlowJo LLC, Ashland, OR).

Neutralization assay
Cd300lf -/or Cd300lf +/mice were infected PO with 10 6 PFU MNoV CW3 and terminally bled via cardiac puncture at 14 dpi. Whole blood was collected in an EDTA Microtainer tube (BD) and pelleted at 5800 x g for 10 min. Plasma was removed and stored at 4˚C until use. BV2 cells were seeded at 20,000 cells in 50μL per well in a 96-cell plate and incubated at 37˚C. Sera were serially diluted in D10 with an initial dilution 1:9 and with five subsequent three-fold dilutions. Then, 10 5 PFU of MNoV CW3 was added to each well and mixtures were gently rocked for 30 min at room temperature (RT) and then added to BV2 cells. After 48 hours, 25μL of CellTiter-Glo was added to each well and luminescence was detected on a luminometer. Experiments were performed in duplicate with at least three independent replicates.

MNoV-specific ELISA
In a 96-well MaxiSorp plate, 100μL of two-fold serially diluted IgG (starting at 12.5ng/mL) and IgM (starting at 25ng/mL) was used as standard controls and MNoV CR6 was used to coat the plate overnight at 4˚C. The plate was then washed three times with 300μL wash buffer (0.05% Tween-20 in PBS) and then blocked with 100μL of blocking solution (1% BSA in PBS) for 1 hour at RT. After washing 3 times, 50μL of sample at appropriate dilutions in blocking solution was added to each well and incubated for 2 hours at RT. The plate was further washed three times and 50μL of anti-mouse IgG-HRP (A3673, Sigma-Aldrich) or anti-mouse IgM-HRP (A8786, Sigma-Aldrich) diluted in blocking solution was added to each well, then incubated for 2 hours at RT. Wells were washed three times, and then 40μL of ELISA TMB substrate solution (eBioscience, San Diego, CA) was added to each. Plates were incubated for 20 min, then 20μL of stop solution (2N H 2 SO 4 ) added and OD determined at 450 nm and reference wavelength 570 nm.

ELISA for HuNoV VLPs binding to PGM
Pig gastric mucin (Sigma-Aldrich) was immobilized at 10 μg/ml in PBS on 96-well-plate at 4˚C overnight. Wells were washed with PBS-0.05% Tween 20 (PBS-T) three times. After blocking with 5% non-fat milk at RT for 1 hour, plates were washed once with PBS-T. Fc-huCD300 proteins (10 μg/ml) were mixed with purified HuNoV VLPs at indicated concentrations and then added and incubated for 1 hour at RT. The wells were washed three times with PBS-T followed by addition of rabbit anti-VP1 sera diluted in PBS (1:2000). Wells were incubated at RT for 1 hour, followed by washing with PBS-T three times as described previously [44]. Secondary goat anti-rabbit-HRP diluted in PBS were added and incubated for 1 hour. After five washes, TMB substrate was added, and 2M H 2 SO 4 was applied to stop the reaction. Absorbance at 450 nm was measured.

Statistical analysis
All statistical analysis was performed in Prism GraphPad version 8 (San Diego, CA). Error bars represent the standard error of the mean unless otherwise indicated. Mann-Whitney tests were performed for all non-normally distributed data whereas normally distributed data was analyzed using Student's T-tests. All statistical tests were two-sided. Virus heat inactivation was analyzed using a one-way ANOVA. Survival experiments were analyzed by Kaplan-Meier
Supporting information S1 Fig. Tree and VP1 alignment showing CD300 interacting sites. VP1 is the major structural protein of MNoV and is comprised of a shell and protruding domain. The complete VP1 sequence of MNoV strains CW3, CR6, CR3, CR7, WU23, MNV3, and S99 were aligned. The VP1 shell domain comprises the core of the virion and is sufficient for virion assembly. The protruding domain mediates binding to CD300lf and bile salts and is comprised of discontinuous P1 and P2 subdomains. The CD300lf and secondary bile acid (GCDCA) binding sites are highlighted as is the F514I mutation which emerged during infection of Cd300l +/-Stat1 -/and Cd300lf -/-Stat1 -/mice 13 . Secondary structures labeled as alpha-helices (α), 3 10 -helices (η), and beta-strands (β). The number following the annotation is the numerical order of that secondary structure. Helices are displayed as squiggles and strands are represented by a forward moving arrow under the annotation. TT = strict β-turns and TTT = strict α turns. (TIF)