Amino acid substitutions in norovirus VP1 dictate cell tropism via an attachment process dependent on membrane mobility

Viruses interact with receptors on the cell surface to initiate and co-ordinate infection. The distribution of receptors on host cells can be a key determinant of viral tropism and host infection. Unravelling the complex nature of virus-receptor interactions is, therefore, of fundamental importance to understanding viral pathogenesis. Noroviruses are non-enveloped, icosahedral, positive-sense RNA viruses of global importance to human health, with no approved vaccine or antiviral agent available. Here we use murine norovirus as a model for the study of molecular mechanisms of virus-receptor interactions. We show that variation at a single amino acid residue in the major viral capsid protein had a key impact on the interaction between virus and receptor. This variation did not affect virion production or virus growth kinetics, but a specific amino acid was rapidly selected through evolution experiments, and significantly improved cellular attachment when infecting immune cells in suspension. However, reducing plasma membrane mobility counteracted this phenotype, providing insight into for the role of membrane fluidity and receptor recruitment in norovirus cellular attachment. When the infectivity of a panel of recombinant viruses with single amino acid variations was compared in vivo, there were significant differences in the distribution of viruses in a murine model, demonstrating a role in cellular tropism in vivo. Overall, these results highlight the importance of lipid rafts and virus-induced receptor recruitment in viral infection, as well as how capsid evolution can greatly influence cellular tropism, within-host spread and pathogenicity.

detection for TCID 50 assay. Data shows mean TCID 50 /mL (n = 3 ± SEM). . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 17, 2023. ;https://doi.org/10.1101https://doi.org/10. /2023 1 1 being allowed to adhere to the culture vessels ( Figure 2C). In this setup, the titre of 2 1 4 the MNV-1.CW1 I301 variant was again significantly higher than all other infectious 2 1 5 clones, except MNV-1.CW1 S301, with both titres ~10-fold greater than for MNV-2 1 6 1.CW1 T301, V301, L301 and D301 variants. Once again, MNV-1.CW1 P301 and 2 1 7 K301 viral recovery was at or below the LOD. To rule out differences in transfection efficiency, we conducted similar experiments 2 1 9 whereby select infectious clone RNA was co-transfected into BHK-21 cells alongside  Once again, MNV-1.CW1 I301 had a significantly greater viral titre compared to all 2 2 6 other infectious clones when the TCID 50 assay was conducted in BV-2S cells 2 2 7 (Supplemental Figure 1A). There was no significant difference in viral titres between 2 2 8 infectious clones in adherent BV-2 cells (Supplemental Figure 1B). Although there 2 2 9 was no significant difference in adherent BV-2 cells infected in suspension 2 3 0 (Supplemental Figure 1C), there was a similar pattern to BV-2S cells across the 2 3 1 infectious clones. Taken together, our data suggest that the MNV-1.CW1 I301 variant has a selective 2 3 3 advantage at infecting cells when in suspension, but no selective advantage is 2 3 4 observed in adherent cells. To determine whether the differences between MNV-2 3 5 1.CW1 I301 and MNV-1.CW1 T301 viruses applied to another cell type, cell culture infectivity assays were performed in the suspension-grown mouse B lymphocyte cell 2 3 7 line WEHI-231. Cells were infected as before and an MTS assay was used to 2 3 8 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made be performed as WEHI-231 cells do not adhere to tissue culture plates; Figure 2D).

4 0
The viral titre of MNV-1.CW1 I301 was ~5-fold greater compared to MNV-1.CW1 2 4 1 T301. In comparison, an infectious clone carrying a replication-defective mutation in 2 4 2 the viral polymerase (GNN) (35) had viral recovery below the LOD. One possible explanation for our observations is that some of the VP1 301 variations 2 4 4 affect virion assembly, not receptor engagement. To investigate this, the total 2 4 5 production of viral particles was measured by one step RT-qPCR. Virus was 2 4 6 produced from infectious clone RNA by transfection into BHK-21 cells as before, and 2 4 7 non-encapsidated RNA was degraded by nuclease treatment before RNA was 2 4 8 extracted and the protected RNA concentration measured by one-step RT-qPCR 2 4 9 ( Figure 2E). There were no statistically significant differences in the number of virus 2 5 0 particles produced by any of the clones. Taken together, these data indicate that the VP1 301 residue plays an important role 2 5 2 in MNV infection, but that this is unrelated to viral replication.

5 3
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made  control. Data shows mean TCID 50 /mL, with significance compared to I301 using one- . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made concentration was then measured by one-step RT-qPCR. Data show mean RNA 2 6 7 genome copies/mL (n = 3 ± SEM).

6 8
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The amino acid at VP1 301 affects cell attachment 2 6 9 VP1 residue 301 contributes to the virus-CD300lf receptor interface (25) and our 2 7 0 data suggested that variation in this amino acid alone is sufficient to confer a 2 7 1 replicative advantage to the virus. We hypothesised that the VP1 I301 variant has residues.

7 7
To prevent endocytosis, binding assays were conducted on BV-2S cells treated with 2 7 8 dynasore (Ds), an inhibitor of dynamin that is required for MNV internalisation (36). BV-2S cells were pre-treated with Ds at 37 C, or left untreated as a control, prior to 2 8 0 incubation with MNV-1.CW1 I301 or T301 viruses. The amount of virus attached to 2 8 1 the cells was measured by western blot for the major viral structural protein, VP1.

8 2
When analysed by western blot ( Figure 3A) and normalised to GAPDH expression, 2 8 3 significantly more MNV-1.CW1 I301 binding was detected compared to MNV-1.CW1 2 8 4 T301 in the Ds pre-treated cells ( Figure 3B). There was also a trend of increased 2 8 5 binding of MNV-1.CW1 I301 compared to MNV-1.CW1 T301 in untreated cells, but 2 8 6 this was not statistically significant. BSA was used as a loading control for 2 8 7 supernatant, due to the presence of FCS (which contains BSA) in the cell media.  I+D  T+D  I  T  I+D  T+D  I  T   I  +Ds   T  +Ds   I  T   Bound MNV  Supernatant   I  +Ds   T  +Ds   I  T . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made Taken together, our observations suggest that viruses encoding I301 have a 3 0 2 selective growth advantage and increased cell attachment when in suspension. To 3 0 3 confirm that differences between cell types were not due to differences in replication 3 0 4 rates, a one-step growth curve with MNV-1.CW1 T301 (10 PFU/cell) was carried out  MNV binding to the cell. We therefore first reduced membrane mobility by performing 3 2 2 the binding assay at a reduced temperature. BV-2S cells were pre-treated with Ds at 3 2 3 37 C (to inhibit internalisation), prior to incubation with MNV-1.CW1 I301 or T301 3 2 4 viruses at either 0 C or 37 C. At reduced temperature there was significantly less 3 2 5 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made 4A & 4B). Next, we treated cells with methyl-β-cyclodextrin (MβCD) and losartan 3 2 7 (Los), two compounds reported to chemically restrict membrane mobility (41, 42). To  for multiple comparisons (n = 3 ± SEM, *p<0.05; **p<0.01.) 3 5 5 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The amino acid at VP1 301 influences MNV infectivity and tropism in vivo
Our data suggested that MNV-1.CW1 I301 infected suspension cells more effectively 3 5 7 than viruses with other amino acids at this position. We therefore hypothesised that 3 5 8 this variation at VP1 301 may affect the cellular tropism in a murine model, due to  the IL in comparison to MNV-1.CW1 I301, but these were not statistically significant.  weight (in g). Data show mean PFU/g with significance compared to I301 using two-3 8 6 way ANOVA with corrections for multiple comparisons (n = 5 ± SEM, *p<0.05). with our findings demonstrating that MNV-1.CW1 I301 has increased cellular tropism 3 9 7 in the spleen. Together, this work reveals the significance of the VP1 301 residue in 3 9 8 MNV infectivity and pathogenesis.  The binding of MNV to CD300lf is thought to be a low affinity, high avidity interaction 4 1 2 that requires a network of hydrophilic and hydrophobic interactions (25). Each virion 4 1 3 is thought to interact with multiple CD300lf receptors, forming clusters that increase 4 1 4 avidity (25, 44). Our results are consistent with this hypothesis and also suggest that  The results described in our study suggests that I301 plays a physiologically 4 3 4 important role. It must be noted, however, that a previous study suggested the T301I 4 3 5 substitution is a tissue culture adaptation, with MNV-3 collected from mice faeces 56 4 3 6 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 17, 2023. ;https://doi.org/10.1101https://doi.org/10. /2023 2 5 days post-infection reverting to T301 (15), which we did not observe in our of HNV infectivity.
CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 17, 2023. ;https://doi.org/10.1101https://doi.org/10. /2023   The plaque assay was performed from virus isolated from mouse tissue as 5 0 7

MTS assay
previously described (59, 63). Data were normalized to the tissue weight and  . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made amplified by RT-PCR using Superscript IV (Invitrogen), followed by second strand  One-step RT-qPCR 5 1 7 Virus sample was treated with 25 U/mL recombinant HS-Nuclease (MoBiTec) at Step RT-qPCR System (Promega), using established primers (31). CT values were 5 2 1 converted to RNA copies/mL by analysing against a pT7-MNV* RNA standard curve  SDS-PAGE and western blot analysis was carried out as previously described (27).

3 1
Blots were analysed on the G:BOX Chemi XX6 machine (Syngene). . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made Detached adherent BV-2 cells or BV-2S cells (2x10 6 /mL) were analysed for CD300lf 5 3 4 expression using a flow cytometry protocol previously described (64), with anti-5 3 5 CD300lf primary antibody (MAB27881, R&D Systems) and Alexafluor647 goat Anti-  Viral binding affinity to BV-S cells was determined using a binding assay modified 5 4 0 from Berry and Tse (33). 10 5 BV-2S cells were pre-treated with 50 µM dynasore (Ds; for 30 minutes at 37 C. MNV was added to the cells at an MOI of 1 and incubated at 5 4 3 either 0 C or 37 C for 2 hours, before completing the binding assay as described.  indicated by *p < 0.05; **p < 0.01; ***p < 0.001.

4 9
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 17, 2023.  Capsid Binding to Bile Acids. J Virol 93:e01581-18.   . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 17, 2023. ;https://doi.org/10.1101https://doi.org/10. /2023