Evolution of SARS-CoV-2 in the murine central nervous system drives viral diversification

Abstract

Severe coronavirus disease 2019 and post-acute sequelae of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection are associated with neurological complications that may be linked to direct infection of the central nervous system (CNS), but the selective pressures ruling neuroinvasion are poorly defined. Here we assessed SARS-CoV-2 evolution in the lung versus CNS of infected mice. Higher levels of viral divergence were observed in the CNS than the lung after intranasal challenge with a high frequency of mutations in the spike furin cleavage site (FCS). Deletion of the FCS significantly attenuated virulence after intranasal challenge, with lower viral titres and decreased morbidity compared with the wild-type virus. Intracranial inoculation of the FCS-deleted virus, however, was sufficient to restore virulence. After intracranial inoculation, both viruses established infection in the lung, but dissemination from the CNS to the lung required the intact FCS. Cumulatively, these data suggest a critical role for the FCS in determining SARS-CoV-2 tropism and compartmentalization.

Figure 1 |. Viral RNA from the CNS has a higher degree of genetic divergence regardless of vaccination status.

a, K18-hACE2 mice were vaccinated with Ad5 vaccines encoding spike, nucleocapsid, spike + nucleocapsid, or PBS as a negative control (n = 5 per group) three weeks before intranasal challenge with SARS-CoV-2. Brain and lung tissue were harvested 5 dpi. b, Phylogenetic tree of the consensus SARS-CoV-2 whole genome sequences (red tips = brain isolate, light blue tips = lung isolate). The shape of the node represents the vaccine given to each mouse, as indicated in the panel legend. Branch length reflects nucleotide substitutions per site as indicated by the scale bar. c, Shannon entropy per position in the brains (red) and lungs (blue) of the indicated mice, separated by vaccine type. N=20 animals (5 per vaccine) were analyzed resulting in a total o 1,127 variable nucleotide positions included from brain and 1,476 positions from lung. Statistics reflect the FDR-adjusted p-values of pairwise comparisons within model fitted after controlling for animal and position in the genome (log(Sh) ~ Tissue* Vaccine + (1 | Animal) + (1 | POS)). Box plots represent the median and interquartile range. d, Shannon entropy per position across the genome for the viral inoculate (n = 1, lower, dark blue), the lung isolates (n = 20, middle, light blue), and the brain isolates (n = 20, upper, red). The coding region for the spike protein is back shaded in gray with the FCS position labelled. e, Alignment of each nucleotide consensus sequence across the Spike FCS region.

Figure 2 |. BALB/c mice infected with mouse-adapted SARS-CoV-2 have increased viral diversity in the CNS.

a, BALB/c mice were intranasally challenged with mouse-adapted SARS-CoV-2 (n = 5 mice). Brain and lung tissue were harvested 5 dpi. (n=5). b, Ct values of the SARS-CoV-2 viral RNA (N1 target) isolated from the brain and lung of each mouse as quantified by RT-qPCR. The black line indicates the average Ct value per compartment. c, Phylogenetic tree of the consensus SARS-CoV-2 whole genome sequences (dark blue tip = inoculate, red tips = brain isolate, light blue tips = lung isolate). Branch length reflects nucleotide substitutions per site. d, Alignment of each consensus sequence across the Spike FCS region. e, Box plots of the Shannon entropy at each position across the genome for the brain isolates (n = 5, red), the lung isolates (n = 5, light blue), and the viral inoculate (n = 1, dark blue). Box plots represent the median and interquartile range. A log-transformed linear mixed effects model was used to test for significant differences in overall genetic entropy between tissues (p-value = 4.43×10⁻¹²). f, Shannon entropy per position across the genome for the viral inoculate (n = 1, lower, dark blue), the lung isolates (n = 5, middle, light blue), and the brain isolates (n = 5, upper, red).

Figure 3 |. Mutation of the FCS blocks Spike S1/S2 cleavage and mitigates utilization of an alternative entry pathway.

a, Pseudoviruses encoding for luciferase were generated with the WT or ΔFCS deletion version of spike. Protein was isolated from pseudovirus cultures and subject to immunoblot staining against the spike protein. Shown is a representative blot of three independent western blots. Quantification of band intensity represented with density of Full Length/density of S1-S2. b, Vero-E6 cells expressing hACE2 and TMPRSS2 were challenged with WT or ΔFCS pseudotyped virus in the presence of aloxistatin or camostat mesylate at increasing concentrations. After 48h of infection, cells were lysed, and luciferase expression was quantified. Relative luciferase expression was normalized to untreated cells. The data are presented as the mean percentages of inhibition across six wells. P-values comparing curve fit of WT and ΔFCS viruses are reported (extra sum-of-squares F test, error bars on points represent mean +/− SD). c, Luciferase expression of ΔFCS or WT pseudovirus in Calu3 cells over luciferase expression in Vero-E6 expressing hACE2 and TMPRSS2 (p-value two-tailed unpaired t-test). d, WT and ΔFCS virus were derived from infectious clones. Protein was isolated from viral stocks and subject to immunoblot staining against the spike protein. Shown is a representative blot of three independent western blots. Quantification of band intensity represented with density of Full Length/density of S1-S2.

Figure 4 |. ΔFCS virus is attenuated after intranasal infection.

a, K18-hACE2 mice were intranasally inoculated with WA-1 and ΔFCS stocks. Mice were intranasally challenged with 6×10³ PFU of wild-type (red/purple) or ΔFCS (blue/green) SARS-CoV-2 and b, evaluated for weight loss (n = 15 for each virus, shown is the mean +/− SEM). At 2 days post-infection (dpi), n = 5 mice per virus were euthanized with remaining n = 10 euthanized at 5-dpi. Viral genomes/organ isolated from the c, lung and f, brain were quantified via RT-qPCR and plotted. Infectious virus was determined using focus forming assay in a 96 well plate with VeroE6 cells using d, lung and g, brain homogenate. Statistical significance compared to WT virus using two-tailed unpaired T test. e, Lungs of infected mice at 2 or 5 days post infection were harvested, fixed, and paraffin embedded. Immunohistochemistry staining for the nucleocapsid protein was performed and images collected at 4x magnification. h, Brains of infected mice at 5 days post infection (n = 3 per virus, representative image shown) were stained for the nucleocapsid protein via immunohistochemistry and images collected at 20x magnification.

Figure 5 |. ΔFCS virus replicates faster in the CNS after intracranial inoculation.

a, K18-hACE2 mice were intracranially challenged with 10² PFU of wild-type (red/purple) or ΔFCS (blue/green) SARS-CoV-2. b, Mice were evaluated for weight loss (n = 15 for each virus, shown is the mean +/− SEM). At 1 day post-infection (dpi), n = 10 mice per virus were euthanized with remaining n = 5 euthanized at 3-dpi. Viral RNA from the c, brain and f, lung was isolated and quantified via RT-qPCR with viral genomes/organ plotted. Infectious virus was determined using focus forming assay on VeroE6 cell with d, brain and g, lung homogenate. Statistical significance is noted if p-value < 0.05 as compared to WT virus (two-tailed unpaired T test). No asterisk signifies no statistically significant difference. e, Brains of infected mice at 1 day post infection were harvested, fixed, and paraffin embedded. Immunohistochemistry staining for the nucleocapsid protein was performed and images of the corpus callosum collected at 10x magnification. h, Lung homogenate at 3-dpi from WT or dFCS infected mice were inoculated onto a monolayer of Vero-E6 cells and incubated for three days. As a control, lung homogenate was UV treated to inactivate infectious virus in one group. Brightfield images of the monolayer is shown.

Figure 6 |. There is selective pressure for the deletion of the FCS in the CNS.

a, Phylogenetic tree of the consensus SARS-CoV-2 whole genome sequences after intranasal or intracranial challenge with WT or ΔFCS virus (dark blue tip = inoculate, red tips = intracranial challenge, green tips = intranasal challenge, triangle = brain isolate, square = lung isolate). Branch length reflects nucleotide substitutions per site. b, Alignment of each consensus sequence across the Spike FCS region after challenge with the WT (left) or ΔFCS (right) virus. c, Phylogenetic tree of SARS-CoV-2 quasispecies after intranasal or intracranial challenge with WT (left) or ΔFCS virus (right) (dark blue tip = inoculate, red tips = intracranial challenge, green tips = intranasal challenge, triangle = brain isolate, square = lung isolate). Branch length reflects nucleotide substitutions per site and tip size reflects the frequency of a given subpopulation in an isolate.