Vero E6 cells (ATCC CRL-1586, Manassas, VA, USA) were grown in Dulbecco’s modified Eagle medium (DMEM) supplemented with 5% heat inactivated fetal bovine serum (FBS). SARS-CoV-2, isolate hCoV-19/South Africa/KRISP-EC-K005321/2020, NR-54008, lineage B.1.351 was obtained from BEI Resources, NIAID, NIH (Manassas, VA, USA) and propagated in Vero E6 cells. Authenticity of the viral strain was validated by genome sequencing and compared to published sequence (GISAID Accession ID EPI_ISL_678597). Virus titers were determined by focus-forming assay on Vero E6 cells. All procedures with infectious SARS-CoV-2 were performed in certified biosafety level 3 (BSL3) facilities at the University of Minnesota (UMN) using appropriate standard operating procedures (SOPs) and protective equipment approved by the UMN Institutional Biosafety Committee.

C57BL/6J mice (The Jackson laboratory Stock No. 000664) and K18-hACE2 mice (The Jackson laboratory, Stock No. 034860) were propagated at the UMN and maintained in specific pathogen-free conditions. Mice were housed in groups of 3 to 5 per cage and maintained on a 12 h light/12 h dark cycle with access to water and standard chow diet ad libitum. Both male and female C57BL/6J mice at different ages (4-, 10-, and 16-months) or 4-month-old hemizygous K18-hACE2 mice were used in this study.

Mice were randomly assigned to infected and uninfected groups with equal numbers of male and female mice in each group. Mice in the infected group were anesthetized with 3% isoflurane/1.5 L/min oxygen in an induction chamber and inoculated intranasally with 1 × 10⁵ focus forming unit (FFU) of SARS-CoV-2 in a volume of 50 μL DMEM, split equally between each nostril. Mice were monitored, and their body weight measured daily for the duration of the experiment. At 3- and 7-days post infection (dpi), mice were euthanized, and lungs and brains were harvested for downstream analysis.

Approximately half of the lung and the left hemisphere of each brain were weighed and homogenized in 1 mL of DMEM supplemented with 2% FBS in GentleMACS™ M tube using GentleMACS™ Dissociator (Miltenyi Biotec) with RNA 2.01 program setting. Tissue homogenates were stored in aliquots at −80°C until use. Total RNA was extracted from 250 μL of tissue homogenate using Trizol™ LS reagent (Thermo Fisher scientific, Waltham, MA) according to the manufacturer’s instructions. The purity of RNA was assessed by calculating the ratio of OD₂₆₀/OD₂₈₀. 1 μg of total RNA was reverse transcribed using High-Capacity cDNA reverse transcription kit (Thermo Fisher scientific, Waltham, MA) according to manufacturer’s instructions. RT-qPCR was performed using Fast SYBR Green master mix (Thermo Fisher scientific, Waltham, MA) in a 7500 Fast Real-time PCR system (Applied Biosystems) using the Centers for Disease Control and Prevention RT-PCR primer set targeting SARS-CoV-2 N gene, forward primer 5′-GACCCCAAAATCAGCGAAAT-3′ (2019-nCoV_N1-F), reverse primer 5′-TCTGGTTACTGCCAGTTGAATCTG-3′ (2019-nCoV_N1-R). The following reaction conditions were used: 95°C for 3 min followed by 40 cycles of 95°C for 10 s, 65.5°C for 30 s, and 72°C for 30 s. A standard curve was generated to determine genome copy numbers in the tissue sample by performing a RT-qPCR using synthetic SARS-CoV-2 RNA control (Twist Bioscience, Cat #104043). Viral loads are presented as copies per μg of total RNA.

Vero E6 cells were seeded in 24-well tissue culture plates at a density of 1 × 10⁵ cells/well and incubated until the monolayer was 90–100% confluent. Four replicate wells of Vero E6 cells were infected with 100 μL of 10-fold serial dilution of virus or lung/brain homogenate and incubated for 1h with intermittent mixing at 37°C/ 5% CO₂. After 1 h, 500 μL of overlay medium containing 1.6% microcrystalline cellulose, 2% heat-inactivated FBS in DMEM was added to each well and incubated for 48 h at 37°C/ 5% CO₂. The cells were fixed with 4% paraformaldehyde for 30 min at room temperature. Fixed cells were washed with PBS containing 0.3% Triton X-100 (PBST), blocked with blocking buffer (1% bovine serum albumin (BSA) and 1% normal goat serum in PBST) for 1 h at room temperature, and then treated with rabbit anti- SARS-CoV-2 nucleocapsid antibody (Sino Biological; 1:2000 dilution) overnight at 4°C. After washing twice with PBST, cells were treated with alkaline phosphatase-conjugated goat anti-rabbit IgG (Thermo Fisher scientific, Waltham, MA; 1:1000 dilution) at room temperature for 1 h. Cells were washed twice with PBST and incubated for 20 min at room temperature in the dark with 1-Step™ NBT/BCIP substrate solution (Thermo Fisher scientific, Waltham, MA). After incubation, wells were washed with deionized water and the numbers of foci were counted. Infectious virus titer in tissue homogenate was expressed as focus forming units (FFU) per g tissue.

Lung samples were immersion-fixed with 4% paraformaldehyde (PFA) for 48 h and washed 3 times with PBS. Samples were routinely processed and embedded in paraffin. From each paraffin block, which contained representative tissue sections from 3 animals, 5 micron sections were cut and stained with hematoxylin and eosin (H&E). Each section was evaluated by a board-certified veterinary pathologist to provide a semi-quantitative assessment of lung injury and a descriptive narrative of lung pathology. The lung injury parameters (perivascular inflammation, interstitial inflammation, interstitial / perivascular edema, alveolar edema, thrombosis, and hemorrhage) were scored (in 0.5 increments) on a 5-point injury scale of increasing severity and extent of tissue involvement. Inflammation was further characterized according to the nature of the cellular infiltrate (i.e., neutrophilic, eosinophilic, mononuclear, histiocytic, or mixed).

Mouse brain hemispheres were immersion-fixed with 4% PFA for 48 h and washed 3 times with 1X PBS. Brains were embedded in Tissue-Tek OCT (Andwin Scientific, IL) on dry ice. Frozen blocks were stored at −80°C until sectioning. Embedded brains were cut into sagittal sections at 10 μm thickness using a Cryostat (Leica Biosystems Inc). Sections were washed with wash buffer (0.3% Triton-X 100 in 1X PBS) and blocked with blocking buffer (1% normal goat serum, 0.3% Triton-X 100 in 1X PBS) for 60 min at room temperature. Tissue sections were incubated overnight at 4°C with primary antibody against the SARS-CoV-2 nucleocapsid protein (1:1000, Sino Biologicals US Inc, Wayne, PA). After washing 3 times with wash buffer, sections were incubated with goat anti-rabbit IgG-Alexa Fluor™ 594 (1:1000, Thermo Fisher Scientific, Waltham, MA) for 60 min at room temperature. DAPI (4′,6-diamidino-2-phenylindole, Thermo Fisher Scientific, Waltham, MA) was used to label nuclei. Images were captured using Leica DMi8 inverted microscope and Leica LAS software.

RNA was extracted from homogenized tissue as described above. Isolated RNA samples were submitted to the UMN Genomics Center core facility for library preparation and sequencing as previously described (Jeong et al., 2021; Qu et al., 2023). Briefly, RNA quantities were determined by RiboGreen RNA Quantification assay (Thermo Fisher Scientific, Waltham, MA) and RNA quality was assessed using the Agilent Bioanalyzer (Agilent Technologies). Library preparation was performed using a TruSeq Stranded mRNA Library Prep Kit (Illumina) according to the manufacturer’s instructions. Sequencing was performed on a NovaSeq 6000 platform (Illumina), generating 20 million 150-bp paired-end reads per sample.

Raw sequencing reads were evaluated and trimmed using Trimmomatic (Bolger et al., 2014). Trimmed reads were mapped to the mouse reference genome GRCm38 using HISAT2.¹ RNA-seq raw counts for each gene were extracted using FeatureCounts (Liao et al., 2014). Exploratory data analysis using PCA was performed on normalized data for outlier detection. One outlier (uninfected, male) was detected and excluded from subsequent analyses (Supplementary Figure 1). Differential gene expression (DGE) analysis was conducted using DESeq2 (1.38.2) (Love et al., 2014). Default DESeq2 normalization and filtering methods were applied.

Data from both sexes (male and female) were analyzed in pairwise comparisons between infected and uninfected males and females, in addition to pairwise comparisons between infected and uninfected mice using the Wald test (design = ∼ sex + infection). Additionally, to assess potential differences in the infection effect between sexes, we performed analyses using design = ∼ sex + infection + sex:infection. Significance for gene differences between groups was determined using an adjusted p-value of <0.05 after Benjamini-Hochberg correction. GO gene enrichment analysis was performed using the gseGO function in the ClusterProfiler package (4.6.0) (Yu et al., 2012). Volcano plots were generated using ggplot2 (3.4.0). Hierarchical cluster analysis of the differentially expressed genes (DEGs) was performed using pheatmap (1.0.12) with default parameters (clustering_distance_cols = euclidean, clustering_method = complete). All DEG and pathway enrichment analyses were performed in R (4.2.1).

Total RNA was extracted from lung and brain homogenates and cDNA was synthesized from 1 μg total RNA as described above. The cDNA was amplified by RT-qPCR using Fast SYBR Green master mix (Thermo Fisher scientific, Waltham, MA) in a 7500 Fast Real-time PCR system (Applied Biosystems) using gene specific primers (Table 1). The specificity of RT-qPCR was assessed by analyzing the melting curves of PCR products. Ct values were normalized to RPL27 gene, and the fold change was determined by comparing virus-infected mice to uninfected controls using 2^–ΔΔCt method (Livak and Schmittgen, 2001).

Statistical analyses were performed using Prism version 9.5.1 (GraphPad, La Jolla, CA). A two-way Analysis of Variance (ANOVA) with Sidak’s multiple comparison test was used to determine the significance of the difference in body weight changes. A two-way ANOVA with Tukey’s multiple comparison test was used to analyze viral load and relative gene expression of RT-qPCR data. A p-value of less than 0.05 was considered significant.

Previous studies have shown that SARS-CoV-2 B.1.351 and other variants with the N501Y mutation in the viral spike protein efficiently infected wild type laboratory mice (Montagutelli et al., 2021; Pan et al., 2021; Shuai et al., 2021; Yasui et al., 2022; Zhang et al., 2022). To determine whether different age groups showed differences in susceptibility to SARS-CoV-2 infection, 4-, 10-, and 16-month-old wild type C57BL/6J mice were infected intranasally with 1 × 10⁵ FFU of SARS-CoV-2 B.1.351 variant (South Africa/KRISP-EC-K005321/20204) and monitored for 7 days, with uninfected age/sex-matched mice as controls. No significant loss of body weight was observed in 4-month-old mice infected with SARS-CoV-2 compared to uninfected mice. However, body weight loss was significantly greater in infected, older male mice at 10- and 16-month of age, than in female mice of the same age (Figure 1). Loss of body weight peaked at 4 dpi in older male mice (male 6.45 ± 1.15% vs female 0 ± 1.49% in 10-month-old; p < 0.01 and male 3.7 ± 1.01% vs female 0.25 ± 1.18% in 16-month-old; p < 0.05), whereas female mice showed no significant loss in body weight compared to uninfected animals. Further loss in body weight was not observed in male mice after 4 dpi and all mice survived until the experimental end point of 7 dpi.

The level of viral RNA and infectious virus titer in the lung homogenate of mice were assessed at 3 dpi and 7 dpi. As shown in Figure 2A, viral RNA was detected by RT-qPCR in the lungs of all SARS-CoV-2 infected mice at 3 dpi. Viral RNA was also detectable in most of the lungs of SARS-CoV-2 infected mice at 7 dpi except for one 4-month-old and one 16-month-old mouse. However, the viral RNA load in the lung at 7 dpi was lower compared to 3 dpi. Significantly higher copies of viral RNA were detected at 3 dpi in the lungs of 16-month-old mice compared to 4-month-old mice (Figure 2A), albeit there was no difference in viral RNA levels between different age groups at 7 dpi. Analysis of data based on sex showed that the older male mice at 16 months of age had significantly higher levels of viral RNA in the lung at 3 dpi compared to 4-month-old male mice (Figure 2B). Of note, the viral RNA load in the lungs of female mice was lower than that in male mice at comparable ages at 3 dpi, although the difference did not reach statistical significance except in 16-month-old mice (p = 0.036) (Figure 2B). Infectious viral load in the lung was measured by focus forming assay on Vero E6 cells. Although the infectious virus was detected in lungs of SARS-CoV-2 infected mice of all age groups at 3 dpi, infectious virus was infrequent in the lung at 7 dpi (Figure 2C). Additionally, lung virus load was significantly higher at 3 dpi in 10-month-old mice compared to 4-month-old mice. At 7 dpi infectious virus was occasionally detected in the lung of younger mice with 14% of 4-month-old (1 out of 7) and 12.5% of 10-month-old (1 out of 8) mice having detectable virus. In contrast, 75% (6 out of 8) of 16-month-old mice harbored detectable levels of infectious virus in the lung at 7 dpi (Figure 2C), suggesting that the virus clearance is slower in older mice. Analysis of sex-disaggregated data showed that 10-month-old male mice had significantly higher infectious viral load in the lung at 3 dpi compared to 4-month-old male mice. Viral load in the lung of female mice at 3 dpi were lower than male mice in all age groups studied, although the difference was statistically significant only in 10-month-old mice (Figure 2D). At 7 dpi, 25% male (1 out of 4) 4-month-old and 25% female (1 out of 4) 10-month-old mice had detectable virus. While all the 16-month-old male mice had infectious virus in the lung, only 50% (2 out of 4) of female mice showed detectable virus in this age group at 7 dpi (Figure 2D). Collectively, these results indicate that older male mice take more time to clear infectious virus from the lung compared to younger mice.

Regardless of age or sex, the lungs from all SARS-CoV-2 infected mice revealed a similar spectrum and, with rare exception, a similar severity of pathology. The infected mice demonstrated multifocal perivascular and interstitial pulmonary inflammation with edema and alveolar histiocytosis (Figure 3 and Supplementary Table 1). In both the male and female mice, the inflammation was most severe in the 10-month-old mice, least severe in the 4-month-old mice, and intermediate in the 16-month-old mice. When considering sex, there was a trend toward more severe inflammation in the female mice compared to male mice at 16 months of age and there was no overall difference between the sexes at 10 and 4 months of age.

SARS-CoV-2 respiratory infection affects other organ systems, including the central nervous system. To investigate the presence of viral RNA in the brain of C57BL/6J mice infected with SARS-CoV-2 B.1.135 variant, RT-qPCR was performed on total brain RNA at 3 dpi and 7 dpi. Viral RNA was occasionally detected in the brain of SARS-CoV-2 infected mice. At 3 dpi one 10-month-old female (1 out of 3) and one 16-month-old male (1 out of 3) had detectable viral RNA in the brain. At 7 dpi very low viral RNA copies (100–150 copies) were detected in two 4-month-old male (2 out of 4), one 10-month-old female (1 out of 4), one 16-month-old male (1 out of 4), and two 16-month-old female (2 out of 4) mice (Figure 4A). Viral RNA was not detectable in the brains of the majority of mice infected with SARS-CoV-2. Moreover, no infectious virus was detected in the brain of C57BL/6J mice in any of the age groups studied (data not shown). In contrast, we found abundant viral RNA and nucleocapsid protein in brains of 4-month-old K18-hACE2 mice infected with 1 × 10⁴ FFU of SARS-CoV-2 (Figures 4B, C), consistent with previous reports (Jiang et al., 2020; Kumari et al., 2021; Song et al., 2021; Fumagalli et al., 2022).

Inflammatory cytokine and chemokine mRNA expression levels at 3 dpi and 7 dpi were measured in lungs and brains by RT-qPCR, including IL-6, TNF-α, and CCL2 (Figure 5 and Supplementary Figure 2). As expected, cytokine expression levels in the lung were higher in infected animals at all age groups (Figure 5A). A higher level of IL-6 mRNA expression was observed in lung of older mice compared to 4-month-old mice. While there was no difference in expression of TNF-α at different age groups, 10-month-old mice showed increased CCL-2 expression compared to 4-month and 16-month-old mice (Figure 5A). Analysis of data based on sex revealed differential expression of IL-6 in male and female mice, which was not consistent with age. A significant upregulation of IL-6 mRNA was observed in the lung of 4-month and 10-month-old male mice at both 3 dpi and 7 dpi, compared to age-matched female mice. In contrast, IL-6 expression was significantly higher in the lung of female 16-month-old mice at 3 dpi (Supplementary Figure 2A). Differential expression of TNF-α and CCL2 mRNA in the lung of male and female mice were observed only in 10-month-old mice. Higher level of expression of TNF-α was observed in 10-month-old female mice at both 3 dpi and 7 dpi. CCL2 expression in the lung of 10-month-old mice was higher in males at 3 dpi in contrast to females at 7 dpi (Supplementary Figure 2A).

In the brain at 3 dpi, IL-6 expression was observed only in 16-month-old mice whereas at 7 dpi, IL-6 expression increased in the brain of infected animals of all age groups compared to uninfected age matched controls (Figure 5B). Interestingly, IL-6 expression level in the brain at 7 dpi decreased with increase in age of infected mice. Although there was upregulation of TNF-α expression in the brain of infected mice at 3 dpi compared to uninfected control, there was no difference in the expression levels between different age groups. Higher CCL-2 expression was observed only in the brain of 10-month-old mice (Figure 5B). There was no significant difference in the level of expression of IL-6, TNF-α, and CCL2 in the brain between male and female mice at 7 dpi. However, there was an age dependent decrease in the expression of IL-6 in the brain in male mice at 7 dpi (Supplementary Figure 2B). A significant increase in IL-6 expression in the brain at 3 dpi was observed only in 16-month-old male mice. Additionally, elevated levels of TNF-α in the brain of 4-month-old female mice and CCL2 in 16-month-old male mice were observed at 3 dpi (Supplementary Figure 2B).

To examine the global effects of SARS-CoV-2 infection on the brain transcriptome, RNA sequencing was performed on brain tissue homogenate collected at 7 dpi from 10-month-old SARS-CoV-2-infected (n = 6) and uninfected C57BL/6J mice (n = 5). Differential gene expression analysis revealed 1067 differentially expressed genes (DEGs). Of these, 478 genes were significantly upregulated (log FC > 1 and adjusted p < 0.05) and 589 genes were significantly downregulated (log FC < 1 and adjusted p < 0.05) in SARS-CoV-2-infected mice (Figure 6A). Hierarchical clustering analysis of significant DEGs (adjusted p-value < 0.05) revealed distinct transcriptional profiles corresponding with SARS-CoV-2 infection (Figure 6B). Differential gene expression analyses aimed at assessing the influence of sex revealed a minimal number of DEGs, suggesting a limited impact of sex on the overall brain transcriptome.

We performed gene set enrichment analysis (GSEA) to identify the major pathways responsible for the differences between groups (Supplementary Table 2). The top upregulated Gene Ontology (GO) biological process terms in SARS-CoV-2 mice were associated with defense response to virus and other organisms. Among these pathways, common genes included Tlr7, Ifit1, Ifit2, Ifih1, and Gbp7, all of which play a role in the innate immune response. The top downregulated GO biological process terms in SARS-CoV-2 mice were associated with neuroreceptor activity, axon development, and cell junction organization (Figures 6C, D). These findings provide a foundation for further investigations into the functional implications of these transcriptomic changes in the context of SARS-CoV-2 infection.

To validate the bulk RNA-seq data, selected immune pathway genes that were upregulated in SARS-CoV-2 infected mice compared to controls (Supplementary Figure 3 and Supplementary Table 3) were tested by RT-qPCR. The mRNA expression levels of Ifit1, Ifit2, Tlr7, Lyz2, B2m, Mpeg1, and Gbp7 were evaluated in 4-, 10-, and 16-month-old mice (Supplementary Figure 4). Consistent with the RNA-seq data, RT-qPCR results showed significant upregulation of Ifit1, Lyz2, and Mpeg1 mRNA in the brain of 10-month-old SARS-CoV-2 infected mice compared to uninfected controls. Although, the expression levels of Ifit2, Tlr7, and B2m were higher than controls, they did not reach statistical significance. There was no change in the expression of Gbp7 compared to uninfected controls. Further analysis of gene expression data from different age groups of male and female mice revealed that there was age-dependent decrease in expression of Ifit1and Lyz2 genes in male mice (Supplementary Figure 4). While no age dependent change was observed for Lyz2 expression in female mice, a non-significant trend toward higher Ifit1 expression levels was observed in older female mice. Interestingly, the expression levels of both Ifit1 and Lyz2 were greater in 16-month-old female mice compared to age-matched male mice. Mpeg1 expression was elevated only in 10-month-old female mice. These results demonstrate that age and sex modify the expression of innate immunity-related genes in the brain in response to SARS-CoV-2 infection.

To dissect specific immune cell subsets from the bulk RNA-seq data, we used CIBERSORTx algorithms to determine immune cell type proportions in each sample (Newman et al., 2019). Among the identified cell types, microglia constituted the majority of cells (SARS-CoV-2 76.2% ± 6.2%, Uninfected 76.0% ± 5.3%), followed by T cells (SARS-CoV-2 15.4% ± 4.9%, Uninfected 19.4% ± 5.6%), macrophages (SARS-CoV-2 7.1% ± 1.2%, Uninfected 3.8% ± 1.7%), monocytes (SARS-CoV-2 1.1% ± 0.43%, Uninfected 0.77% ± 0.46%), and natural killer (NK) cells (SARS-CoV-2 0.28% ± 0.44%, Uninfected 0.050% ± 0.11%) (Figures 7A, B). Intriguingly, the proportion of macrophages was significantly higher in SARS-CoV-2 infected mice compared to uninfected controls (p = 0.0372) (Figure 7C). These findings underscore the impact of SARS-CoV-2 infection on immune cell composition within the brain microenvironment, highlighting the special recruitment of macrophages.

Among the downregulated pathways, GO enrichment analysis showed that pathways related to cell junction organization and regulation of vasculogenesis were altered in SARS-CoV-2-infected brains. Several genes/proteins involved in maintaining the blood-brain barrier (BBB) integrity were downregulated, including Cldn5, a member of the claudin family. Claudins are key components of tight junctions between adjacent endothelial cells that regulate the permeability of the BBB (Morita et al., 1999; Luissint et al., 2012). Multiple proteins with functions related to regulation of endothelial cell-matrix adhesion or cell-cell junction organization (Ajm and Pard6a, Nectin1, Rcc2, Cbln1) were also expressed at significantly lower levels in infected brains. Consistent with the GO analysis observations, we also found a significant decrease in expression of cadherin proteins (Cdh1, Cdh22, Cdh15, Cdh19) in SARS-CoV-2 infected brains. Cadherins are transmembrane proteins that mediate cell–cell adhesion and cell-cell junction organization (Togashi et al., 2009; Maître and Heisenberg, 2013). These results suggest that SARS-CoV2 infection may lead to disruption of BBB integrity and neurovascular function, in addition to triggering the innate immune response in the brain. Further studies are required to experimentally demonstrate BBB permeability after SARS-CoV-2 infection.