VSVs infect olfactory sensory neurons
To evaluate whether VSV gains access to the brain via the olfactory nerve by infecting OSNs, VSV12’GFP was placed at the nostrils of mice and inhaled into the nasal cavity. Temporal and cell-type specific viral transduction were examined by viral GFP expression. No GFP expression was observed in the OE at 6hrs post viral instillation (6 hrPI) (Fig. 1A). Scattered GFP positive cells were first observed at 12 hrPI (Fig. 1B). All GFP positive cells observed at 12 hrPI have the morphology of mature OSNs (Fig. 1B). By 24 hrPI, GFP positive cells are widely distributed within the OE (Fig. 1C). In addition to OSNs, viral GFP expression is also observed in sustentacular cells and basal cells. The OE tissue shows degenerative changes including a decrease in the OE thickness at 48 hrPI with viral GFP expression still evident (Fig. 1D). Viral GFP was observed in the olfactory nerve bundles within the lamina propria at 24 hrPI (Fig. 1C). The VSV-GFP positive olfactory nerve continues into the olfactory nerve layer and glomerular layer of the OB (Fig. 1E-F). Increased GFP expression was observed in the olfactory nerve layer and glomerular layers of the OB at 48 hrPI. The GFP signal also co-localizes with OMP immunoreactivity labeling the olfactory axons and their terminal arbors within the glomeruli of the OB (Fig. 1G-H). No VSV 12’GFP positive OB neurons were observed at either 24 or 48 hrPI [13].
To determine the levels of viral replication in the OM and OB, viral GFP transcript levels were measured by qRT-PCR and compared at 3hr, 6hr, 12hr, 24hr and 48 hrPI. Viral GFP transcripts were identified at 3 hrPI, the earliest time point examined. GFP transcript levels continued to increase in the OM from 6 hrPI to 48 hrPI (Fig. 1I, relative Log10 fold change (LogFC): 3 hrPI 2.34±0.22 ; 6 hrPI 2.75±0.22; 12 hrPI 4.67±0.09; 24 hrPI 6.87±0.19 ; 48 hrPI 7.4±0.007, p<0.05). In the OB, viral gene transcripts were not detected at 12 hrPI (Fig. 1J, LogFC:12 hrPI − 1.44±0.58). Compared to the PBS control, VSV-GFP transcript levels were significantly increased at 24 hrPI and 48 hrPI (Fig. 1J, LogFC: 24 hrPI 4.26±0.15; 48 hrPI 5.4±0.087, p < 0.05).
During the acute phase of viral infection, within 48hrPI, while viral replication is active in OSNs, the rates of apoptosis were investigated. Activated caspase 3 immunohistochemistry was performed on Control and VSV infected OE (Fig. 2). Increased numbers of apoptotic cells were observed compared to the wildtype control in 3hr, 6hr, 12hr, 24hr and 48 hrPI (Fig. 2B, Cells/mm: wildtype 14.2±1.7 3 hrPI 90.6±2.7 ; 6 hrPI 58.3±4.7; 12 hrPI 25.5±1.2; 24 hrPI 35.8±10.8 ; 48 hrPI 100±2.4, p<0.05). Interestingly, increases of apoptosis occurred rapidly at 3 hrPI and subsequently subsided before increasing again at 48 hrPI when OE has undergone pathogenic changes.
Acute changes of transcription profiles in the olfactory mucosa
To better understand innate immune responses in the OM during acute phase of VSV exposure, we conducted RNA-seq analysis to systematically examine gene expression regulation within the first 48 hrPI. Biological triplicates of OM were collected at 3hr, 6hr, 9hr, 24hr and 48 hrPI after VSV or PBS nasal instillation. No significant gene expression changes between VSV and PBS control OM before 24hrPI. At 24 hrPI, 1655 transcripts showed differential expression greater than 2.5 folds in VSV samples compared to PBS (Fig. 3A, adjP < 0.05). The number of upregulated genes was higher than the number of downregulated genes (upregulated genes = 1230 genes; downregulated genes = 425 genes, adjP < 0.05). Amongst the genes showing robust differential expression are cytokines and chemokines (Fig. 3B-C). Gene ontology enrichment analysis was performed. GO terms related to antiviral responses were prominently recognized (Fig. 3D). The expression of selected transcripts was validated by qRT-PCR. Upregulation of Il6, Cxcl10 and Rela at 24hr PI were validated (Fig. 3E-G). Though it did not show significant upregulation at 3hrPI by RNA-seq analysis, significant upregulation of Il6, was detected by qRT-PCR (5.29 ± 0.29, p < 0.05, biological replicates n = 6) (Fig. 3E).
Upregulation of type I and III interferon transcript levels
Gene ontology analysis identified a significant change in cellular response to interferon-beta at 24 hrPI (Fig. 3D). In the mouse genome, there are 14 Ifn α isoforms, 1 Ifnβ isoform (type I) and 2 Ifnλ isoforms (type III). Neither type I nor type III Ifn expression was detected in PBS treated OM. Exposure to VSV induced an upregulation of Ifn α2, α4, α16, β1, λ2 and λ3 at 24 hrPI and 48 hrPI in the OM compared to PBS controls. Changes in transcription of Ifnα2, Ifnα4, Ifnβ1 and Ifnλ2/3 at 12hr, 24hr and 48 hrPI were validated by qRT-PCR (Fig. 4A-D, p < 0.05). Type I and type III IFN receptors are expressed in control OM. Compared to PBS control OM, type I Ifn receptor subunit (Ifnar1) and type III Ifn lambda receptor subunit (Ifnlr1) levels were not changed following VSV exposure at 24 hrPI as determined by RNA-seq analysis (Fig. 4E, adjP < 0.05).
To determine cell type specific expression of Ifn receptors, immunohistochemistry was performed to detect protein expression of IFNAR1 and IFNLR1. Mature OSNs were identified by olfactory marker protein (OMP) immunostaining. IFNAR1 expression was detected throughout the depth of the OE and appears to be localized in the majority of the OM cell types, including OSNs (Fig. 4F). IFNLR1 expression was less ubiquitous in the OE. Immunoreactivity of IFNLR1 in the sustentacular cell body layer at the apical surface of the OE was not evident. However, IFNLR1 expression was detected in OMP positive OSNs throughout the OE (Fig. 4G). IFNLR1 expression was also observed in the olfactory nerve bundles, labeled by OMP immunoreactivity, in the lamina propria. Therefore, IFNLR1 is specifically expressed in mature OSNs.
Activation of interferon signaling in the olfactory epithelium
The activation of type I and type III Ifn receptors results in phosphorylation of STAT1 (pSTAT1) and STAT2 (pSTAT2), which subsequently regulate Ifn stimulated genes (ISGs) expressing to perform antiviral functions [27, 28]. To investigate whether exposure to VSV activates Ifn signaling in the OM, we first examined pSTAT1 and pSTAT2 levels by western blotting. pSTAT1 and pSTAT2 are not detected in PBS control OM while they are clearly present in VSV exposed OM at 24 hrPI (Fig. 5A-B). pSTAT1 and pSTAT2 levels were quantified and normalized against α-tubulin loading control. Comparisons of VSV to PBS in biological triplicate OMs show significant changes in pSTAT1 and pSTAT2 at 24 hrPI (pSTAT1 PBS vs VSV: 0.063±0.018 vs 0.30±0.048; pSTAT2 PBS vs VSV: 0.62±0.4 vs 2.5±1.05, p < 0.05).
To investigate the cell type specific STAT activation, immunocytochemistry was performed to detect pSTAT1 expression in the OE. pSTAT1 was detected in majority of cell types, including OSNs, at low levels in the PBS control. At 24 hrPI VSV, pSTAT1 appears in the nuclei of majority of cell types and robustly in sustentacular cells (Fig. 5C).
Expression levels of ISGs were evaluated in VSV exposed OM and compared with PBS controls by qRT-PCR (Fig. 5D-F). Upregulation of Oas1, Ifit2 and Ifit3 transcript levels were first detected at 12 hrPI (Fold: 12 hrPI: Oas1 2.14±0.075, Ifit2 1.83±0.06, Ifit3 6.5±0.05; 24 hrPI: Oas1 29.8±0.13, Ifit2 16.9±0.11, Ifit3 93.1±0.11 ; 48 hrPI: Oas1 28.49±0.13, Ifit2 23.95±0.14, Ifit3 86.6±0.13 p < 0.05). Increased levels of upregulation of ISGs were observed at 24hr and 48 hrPI compared to that of 12 hrPI.
Interferon signaling is required for suppressing VSV replication in the olfactory mucosa
To determine whether Ifn signaling is required for performing antiviral functions in the OM, we examined VSV viral load in Ifnar1 and Ifnlr1 knockout mice and compared it to Ifnar1 and Ifnlr1 wildtype OM. Relative expression levels of the three viral genes VSV-GFP, VSV-M and VSV-N were measured at 24 hrPI. In Ifnar1 knockout OM, all three viral genes examined are slightly increased relative to the wildtype, but the change is not significant. (Fig. 6A, relative fold Ifnar−/−: VSV-GFP: 1.43±0.042; VSV-M: 1.48±0.03; VSV-N: 1.54±0.052, p > 0.05).
Similarly in Ifnlr knockout OM, slight non-significant increases were observed in VSV viral gene expression compared to wildtype (Fig. 6B, relative fold Ifnlr−/−: VSV-GFP: 1.57±0.09; VSV-M: 1.27±0.12; VSV-N: 1.4±0.07, p > 0.05). However, in OM from Ifnar1−/−/Ifnlr1−/− double knockout mice, significant increases in viral genes expression were detected by qRT-PCR (Fig. 6C, relative fold to PBS: VSV-GFP: 2.4±0.51; VSV-M: 3.5±0.47; VSV-N: 3.43±0.41, p < 0.05). Furthermore, knocking out Stat1 which disrupts both type I and type III Ifn signaling, also results in a significant increase in viral gene expression (Fig. 6D, relative fold: VSV-GFP: 2.87; VSV-M: 5.71; VSV-N: 3.97, p < 0.05). These results indicate that both type I and type III Ifn signaling are required for suppressing viral replication in the OM.
Furthermore, we investigated whether IFNs are sufficient in reducing viral load in the OM. Exogenous IFNβ1, IFNλ2, or PBS was provided via nasal instillation to wildtype mice 1 hour before VSV exposure. Viral transcript levels were examined at 24 hr post viral instillation and compared between PBS and IFN groups in biological triplicates. It was observed that the administration of IFNβ1 significantly decreases the relative number of viral transcripts of all three viral genes in the OM (Fig. 7A, relative fold to PBS: VSV-GFP: 0.18±0.26; VSV-M: 0.21±0.27; VSV-N: 0.19±0.26, p < 0.05). Consistently, viral transcript levels significantly decreased in the OB at 24 hrPI as well when primed with IFNβ1 (Fig. 7B, relative fold to PBS: VSV-GFP: 0.26±0.27; VSV-M: 0.27±0.27; VSV-N: 0.26± 0.27, p < 0.05). When comparing priming with exogenous IFNλ2 to PBS before VSV exposure, significant decreases of viral transcripts were observed at 24 hrPI for all three viral genes in the OM (Fig. 7C, relative fold to PBS: VSV-GFP: 0.17±0.07; VSV-M: 0.12±0.06; VSV-N: 0.17±0.08, p < 0.05). Consistently, viral transcript levels were also significantly decreased 24 hrPI in the OB following IFNλ2 priming of the OE (Fig. 7D, relative fold to PBS: VSV-GFP: 0.2±0.44; VSV-M: 0.29±0.28; VSV-N: 0.65±0.25 p < 0.05). These results indicate that type I and type III IFNs are both sufficient to suppress viral replication in the OM. VSV-GFP expression was examined on OE sections. Consistent with qRT-PCR, VSV-GFP was diminished in IFNβ1 primed OE. GFP positive cells were scarcely scattered in the basal cell layer of the OE in IFNλ2 treated animals (Fig. 7E-G). This data may reflect the expression differences between type I and III receptors.