Characterization of virus neuroinvasion, blood-brain barrier integrity and neuroinflammation following Powassan virus infection in mice Authors: Heather Pathak, Amany Elsharkawy, Emilio E. Espínola, Hussin Rothan, Komal Arora, Chinonye Dim, Shannon Stone, Zainab Nabi, Mukesh Kumar Published: 2026 ABSTRACT Powassan virus (POWV) is a tick-borne Orthoflavivirus transmitted by Ixodes tick species. POWV causes fatal encephalitis in approximately 10-30% of neurological cases, and long-lasting neurological sequelae in approximately 50% of survivors. POWV entry into the central nervous system (CNS) is an important event in determining clinical outcome. In this study, we evaluated viral replication kinetics, neuropathology, as well as host immune response following POWV infection in C57BL/6J (WT) mice. Our data showed that infection with POWV by all inoculation routes, including the intravenous, intraperitoneal, intracranial and subcutaneous, led to severe neuroinvasive disease. We showed that POWV effectively replicates in WT mice, where replication and dissemination resulted in peripheral and neurotropic phases. Viral neuroinvasion correlated with severe neuropathological alterations as well as enhanced blood-brain barrier permeability. Next, we used transcriptomics to compare the induction of effector pathways in the brain during the acute and late stages of POWV infection in mice. At all examined time points, we found several dysregulated genes including genes associated with interferon signaling, neuroinflammation and cell death signaling. We detected significant increase in the protein levels of markers involved in neuroinflammation in POWV-infected brains. Immunofluorescence analyses further validated the transcriptomic findings and demonstrated increased activation of microglia (IBA1) and astrocytes (GFAP), infiltration of peripheral immune cells (CD45), and elevated neuronal cell death (TUNEL) in POWV-infected brains. Increased protein expression of caspase-3 and p16 further indicated activation of apoptotic and senescence-associated pathways. Interestingly, we detected viral RNA and found evidence of neuroinflammation persistence, albeit at lower levels, in mice that survived the acute POWV encephalitis phase. Overall, this study provides a comprehensive understanding of the pathogenic events that occur during the acute and late stages of POWV infection in mice. Introduction Powassan virus (POWV) is a single-stranded, positive-sense RNA (+ssRNA) virus that is closely related to other Orthoflaviviruses, including tick-borne encephalitis virus (TBEV), dengue virus, Japanese encephalitis virus (JEV), and West Nile virus (WNV) (1–3). POWV is the only North American member of the tick-borne encephalitis serogroup of Orthoflaviviruses. POWV is transmitted by Ixodes scapularis ticks, the same vector that also transmits the agent of Lyme disease (4, 5). Transmission of POWV from infected ticks to humans has been documented in the USA, Canada, and Russia. Recent public health records indicate an exponential increase in POWV infections and spread in the northeastern and midwestern United States (6). In humans, POWV is the causative agent of a severe neuroinvasive illness characterized by encephalitis, meningitis, and long-term physical and cognitive disabilities in survivors. The fatality rate can be as high as 30% for infected cases and half of survivors experience persistent neurological deficits for several years. Long-term sequelae include cognitive deficit, spastic hemiplegia, muscle weakness, imbalance, ophthalmoplegia, and memory dysfunction (7, 8). No effective therapies exist for treating individuals with POWV infection, and the pathogenesis of POWV encephalitis is not entirely understood. Neuroinvasion is considered a critical step in the progression of POWV-induced encephalitis. The blood-brain barrier (BBB) is a selective cellular border comprised of specialized cerebral microvascular endothelial cells that protect the central nervous system (CNS) from circulating pathogens and leukocytes (9, 10). Several neurotropic pathogens have evolved strategies to cross the BBB to gain access to the CNS, including direct infection of BBB cells, paracellular entry through the compromised BBB, and transmigration within infected leukocytes via the Trojan horse mechanism (11–13). In addition, inflammatory cytokines such as tumor necrosis factor alpha (TNF-α) and interleukin (IL)-1β have been shown to alter the BBB integrity (14, 15). Neuropathological findings from human POWV cases demonstrate widespread neuroinflammation, neuronal injury, microglial activation, and inflammatory cell infiltration within the CNS (7, 8). Various animal models have been used to study POWV infection. Immunocompetent strains of mice, including BALB/c and C57BL/6, develop severe neurological disease when peripherally inoculated with POWV (16–18). Similar pathological features have also been observed in these animal models, including perivascular inflammation, gliosis, neuronal degeneration, and encephalitis. Although these studies have provided important insights into POWV neuropathogenesis, the precise mechanisms regulating viral neuroinvasion, BBB dysfunction, CNS immune responses, and long-term disease progression remain incompletely understood. In the present study, we analyzed POWV infection in a C57BL6/J (WT) mouse model by different routes of inoculation. We next evaluated BBB permeability and neuropathology in POWV-infected mice in relation to virus replication and neuroinflammation. Additionally, we perform whole-transcriptomic profiling of brain tissues during both acute and chronic stages of infection to identify pathways associated with neuroinflammation, immune activation, and neurological disease progression. Using Ingenuity Pathway Analysis (IPA), we provided a comprehensive transcriptomic analysis that recognizes the gene dysregulation leading to acute disease and subsequent late-stage neurological sequalae following POWV subcutaneous inoculation. This study provides an in-detail description of POWV-induced disease from the peripheries to neurological involvement in mice, shedding light on the pathogenic events that occur in this mouse model of POWV infection. Results Characteristics of POWV infection in C57BL/6J mice To assess whether the route of infection affects POWV-associated disease, we infected mice with 1,000 plaque-forming units (PFU) of POWV LB strain by the intravenous (IV), or intraperitoneal (IP), or intracranial (IC), or subcutaneous (SC) route. Mice were monitored for illness signs and survival for 21 days post infection (dpi). Our data demonstrate that a peripheral POWV infection by all routes leads to severe neuroinvasive disease in mice. Mice infected through IC route rapidly succumbed to infection by 5 dpi. Mice infected through IV and IP routes began losing weight and mortality was first observed at 7 dpi. By 10 dpi all mice infected through both IV and IP routes had met clinical endpoint. Alternatively, mice infected through SC route began to lose significant weight and showed severe symptoms starting at 7 dpi and 30% of the SC-infected mice survived until 21 dpi (Figures 1A, B). All infected mice displayed clinical signs of severe neurological disease, including ruffled fur, malaise, hunched posture, ataxia, tremors, loss of balance and paralysis. Mortality in POWV-infected mice. (A) Kaplan-Meier survival curve was generated for each infectious route (1,000 PFU) and mock groups (n= 4 M and 2 F for mock; n= 5 M and 5 F for subcutaneous (SC); n= 5 M and 5 F for intravenous (IV); n= 4 M and 4 F for intraperitoneal (IP); n= 2 M and 2 F for intracranial (IC)). Statistical significance was determined by Mantel-Cox test. (B) Weight loss for POWV-infected mice (1,000 PFU) with different infection routes and mock-infected mice. Mice were weighed daily for 21 days post infection. Weights are expressed as a percentage of initial body weight prior to infection (day 0). Statistical significance was determined by one-way ANOVA (C) Kaplan-Meier survival curve was generated for POWV-infected mice with different infectious doses and mock groups via SC route (n= 4 M and 4 F for 10,000 PFU group; n= 5 M and 5 F for 1,000 PFU group; n= 6 M and 5 F for 100 PFU group; n= 3 M and 3 F for mock-infected control group). (D) Body weight for POWV-infected mice with different infectious doses and mock groups via SC route. Mice were weighed daily for 21 days post infection. Statistical significance was determined by one-way ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001. Since the SC footpad infection mimics natural infection, we next examined the dose-dependent morbidity and mortality rates in mice following a peripheral SC POWV infection. We inoculated mice with sequentially increasing doses of POWV (100 PFU, 1,000 PFU, and 10,000 PFU). POWV-infected mice showed a dose-dependent morbidity and mortality rates (Figure 1C). A significant difference in survival rates was observed within the POWV infectious doses evaluated by a Log-rank Mantel-Cox test. The lowest infectious dose (100 PFU) showed ~50% mortality which increased to 70% and 100% following infectious doses 1,000 and 10,000 PFU respectively. Mice infected with 10,000 PFU began to lose a significant amount of weight starting at 5 dpi and all mice reached endpoint by 10 dpi. Beginning at day 8 through 14 post infection, 70% of mice infected with 1,000 PFU reached clinical end point. Additionally, about 50% of mice infected with 100 PFU reached endpoint between 10 and 12 dpi (Figure 1D). Notably, mice that survived the acute infection continued to demonstrate mild clinical signs such as limb weakness or walking abnormalities until the termination of the study at 21 dpi. To quantify replication kinetics, mice were inoculated with 1,000 PFU of POWV or PBS (mock) via the SC route and euthanized at various time points post infection to measure the viral titers in the serum and brain. Viral infectivity titers in the serum and brains of POWV-infected mice were measured by plaque assay. Significantly high virus titers were observed in the serum at 2 dpi, increased on day 3 and followed by a rapid decline by 6 dpi (Figure 2A). Low titers of POWV were detected in the brain as early as 3 dpi. Significantly high titers of infectious virus were detected in the brains of all the animals by 6 dpi and remained high through 8 dpi. Low viral titers were detected in the brain of a subset of surviving mice at 21 dpi (Figure 2B). Next, RT-qPCR was performed to quantify viral copy numbers in brains collected at 0, 3, 6, 8, and 21 dpi. PCR data was consistent with plaque data. Interestingly, high levels of viral RNA were detected in the brains until 21 dpi (Figure 2C). Notably, the timing of peak viral titers in the brain coincides with morbidity and mortality in mice. Additionally, mice who survived acute infection showed virus persistence in the brain at 21 dpi. Viral titers in POWV-infected mice. (A) POWV titers in the serum are plotted for each mouse at days 0, 2, 3, and 6 post infection and are expressed as PFU per mL of serum. (B) Infectious viral titers in the brain as determined by plaque assay. POWV titers are plotted for each mouse tissue at days 0, 3, 6, 8, and 21 post-infection and are expressed as PFU per gram of tissue. (C) RNA Viral load in POWV-infected brain tissues was analyzed via RT-qPCR. Viral load data are expressed as genome copies per microgram of RNA after normalization to a standard curve. Data are plotted for each individual mouse. Middle horizontal bars indicate mean values for the group and error bars are SEM. Statistical significance was determined by Kruskal-Wallis test followed by Dunn’s multiple comparison test *p < 0.05; **p < 0.01. (D) Mock- and POWV-infected brain sections stained with dsRNA (green) and DAPI (blue). Scale bars are 150μm and 75μm on the original and enlarged images, respectively. (E) Immunofluorescence imaging of brain sections stained with NeuN (red), dsRNA (green) and DAPI (blue). Representative images are shown (n=4 per group). It is widely accepted that double-stranded RNA (dsRNA), is a replicative intermediate during virus infection and is a trait shared by all viruses (19). It has been reported that dsRNA could be detected by immunofluorescence antibody staining in +ssRNA infections (20). We investigated the presence of dsRNA in mouse brain after peripheral POWV inoculation using immunofluorescence labeling of brain sections harvested at 8 dpi. Robust dsRNA signal was detected mainly in the cerebral cortex and the cerebellum of POWV-infected brains, confirming viral replication in the brain at 8 dpi (Figure 2D). Dual staining with NeuN and dsRNA confirmed viral replication within neuronal cells (Figure 2E). POWV neuroinvasion and impact on neuropathology and BBB permeability in mice. Histopathologic examination of brain sections harvested from POWV-infected mice at 8 dpi revealed significant neuropathology, characterized by perivascular cuffing around blood vessels. Figure 3A shows representative histopathological sections from multiple brain regions, including the cortex, hippocampus, thalamus, and cerebellum, illustrating the distribution and severity of POWV-induced neuropathology across distinct anatomical sites. Perivascular cuffing was consistently observed in all brains examined at this time point and was most pronounced around larger blood vessels, particularly in the cortex and thalamus. Other prominent features included shrunken hyper-eosinophilic neurons and loss of Purkinje neurons in POWV-infected brains (Figure 3A). Neuropathology and BBB permeability in mice. (A) Hematoxylin and eosin (H&E) staining of mock- and POWV-infected brain (8 dpi). Representative histopathological sections from multiple brain regions, including the cortex, hippocampus, thalamus, and cerebellum are shown. Hyper-eosinophilic degenerating neurons (black triangle); perivascular cuffing of immune infiltrates around blood vessels (red triangle); activated microglial cell (blue triangle); infiltration of leptomeninges (green triangle); hyper-eosinophilic Purkinje cell (purple triangle) compared to normal Purkinje cell with homogenized cytoplasm (light blue triangle); area of Purkinje cell loss (orange triangle). (B) BBB permeability was assessed using Evan’s blue dye at days 2, 3, and 6 post POWV infection (n=4 per group). Mice were injected intraperitoneally with 1 ml Evans blue dye (1%), mice were perfused with 1X PBS after one hour. Representative images are shown. We next evaluated BBB permeability at 2, 3, and 6 dpi using Evans blue dye to demonstrate leakage into the CNS. Evans blue is a cationic dye that binds to the albumin present in the serum and forms a complex that cannot pass through the intact BBB. Extravasation of this complex into the brain is indicative of increased BBB permeability (21). Six-week-old male and female WT mice were inoculated with 1,000 PFU of POWV or PBS (mock) SC via the footpad. At 2, 3, and 6 dpi, groups of 4 mice were injected with Evans blue dye intraperitoneally (IP). An hour after Evans blue injection, mice were euthanized and perfused with PBS. No extravasation of dye was observed at day 2 post infection. The dye was detected in the brain at 3 dpi, suggesting an early disruption of the BBB. Levels of dye in the perfused brains increased significantly by 6 dpi (Figure 3B). Overall, these findings indicate that a peripheral POWV infection induces disruption of the BBB in mice. POWV infection induces gene dysregulation and persistent transcriptional changes in the brain. To examine the time-course dynamics of host gene expression induced in the brain following a peripheral POWV inoculation, mice were infected with 1,000 PFU of POWV via SC route. Brains were collected from mock- and POWV-infected mice at 3, 6, 8, and 21 dpi and subjected to whole transcriptome analysis through RNA sequencing (Supplementary Tables 1–4). The number of differentially expressed genes (DEGs) was quantified based on an adjusted p-value (FDR) of < 0.05 and Log2 fold change > |± 1.5| in gene expression. POWV infection induced a significant enrichment of 135, 998, 1303 and 436 genes and a significant downregulation of 5, 55, 172 and 115 genes at 3, 6, 8, and 21 dpi, respectively. There was a progressive increase in the number of upregulated and downregulated DEGs from 3 to 8 dpi. Notably, there was persistent dysregulation of genes at 21 dpi (Figures 4A, B). Next, Ingenuity Pathway Analysis (IPA) software was used to evaluate the top diseases and biofunctions altered by POWV infection (Figure 4C). Interestingly, at 3 dpi we observed early activation of antiviral response as well as activation of microglia. At 6 and 8 dpi, there was significant activation of multiple pathways including chemotaxis, inflammatory response and nervous system neoplasia. Most of these pathways persisted through 21 dpi, albeit at lower levels. Differential gene expression in the brain at different time points following POWV infection. (A) Bar graph showing the number of upregulated DEGs induced in mouse brain after POWV inoculation at days 3, 6, 8, and 21 post infection (n=4 per group). (B) Bar graph showing the number of downregulated DEGs induced in mouse brain after POWV inoculation. DEGs were determined with False discovery rate value (FDR) <0.05 and Log2FC > |±1.5|. (C) Heatmap showing time-dependent comparison analysis of canonical pathways induced in the brain post POWV infection. (D) Heatmap showing time-dependent comparisons of disease and biofunction alterations. (E) Heatmap of time-dependent comparisons of neuronal receptor signaling pathways. Next, we interrogated the top canonical pathways induced in the brain following POWV infection over time based on the activation-z score value. Pathways such as neutrophil degranulation, neuroinflammation signaling, pathogen induced cytokine storm and dendritic cell maturation were among the top activated canonical pathways from 3 to 21 dpi with the highest activation observed at 6 and 8 dpi (Figure 4D). In addition, pathways associated with macrophage activation such as phagosome formation, production of nitric oxide, reactive oxygen species and macrophage classical activation pathway were upregulated with the most intensity shown 6 and 8 dpi. In contrast, pathways such as CTLA4 signaling in cytotoxic T-lymphocytes, PPAR signaling and FXR/RXR signaling pathways were downregulated during POWV infection. Interestingly, the signaling of multiple neuronal receptors including serotonin, oxytocin, opioid and acetylcholine receptors was upregulated at 6 and 8 dpi but were downregulated at 21 dpi in surviving mice (Figure 4E). Robust activation of interferon, chemokines/cytokines and neuronal cell death pathway in POWV-infected brain. During viral infection, pathogen-associated molecular patterns (PAMPS) are recognized by pattern recognition receptors (PRRs) such as toll-like receptors (TLRs), resulting in the activation of Interferon Regulatory factors (IRFs) (22, 23). IRFs are key regulators of type I interferon (IFN) expression which have antiviral and antiproliferative effects to limit viral replication. POWV-infected mouse brain transcriptome sequencing results showed the robust activation of the interferon signaling pathway as well as IFN stimulating genes (ISG) at all time points (Figure 5A). We observed a significant upregulation in several genes including Irf7, Isg15, Usp18, Oas2, and Rsad2 at all time points. Chemokine genes, such as C-X-C motif chemokines (Cxcl9, Cxcl10, and Cxcl13) and C-C motif ligands (Ccl2, Ccl5, Ccl7, and Ccl8), showed strong upregulation, indicating the excessive neuroinflammation and immune cells recruitment in mice brains (Figure 5B). Cytokine genes and their receptors were significantly increased at 6 and 8 dpi. A modest proinflammatory response was observed in the brains at 3 and 21 dpi. Cytokines such as Il6, Il12b, Il10, and several Tnf genes were highly upregulated at 6 and 8 dpi (Figure 5C). Activation of interferon, chemokines/cytokines and cell death pathways in the brain post POWV infection. (A) Dot graph showing the activation Z-score of IFN signaling and heat map showing Log2 fold change (Log2FC) of genes associated with IFN pathway. DEGs are shown at days 3, 6, 8, and 21 post infection. (B) Bar graph showing the Log2 fold change of enriched chemokine genes determined by RNA-Seq in POWV-infected mouse. (C) Bar graph showing the Log2FC of enriched cytokine genes determined by RNA-Seq. (D) Dot graph showing the activation Z-score of cell death related pathways generated by IPA and heat maps showing the Log2FC of enriched genes associated with the death receptor, necroptosis, and pyroptosis signaling pathways. (E) Validation of RNA-seq data was conducted by RT-qPCR from mock and POWV-infected brains at 6, 8, and 21 dpi. Fold change of each gene was first normalized to the GAPDH gene and fold change in POWV-infected mice brain was calculated in comparison to corresponding control mice brain. Log2FC is plotted for each mouse (n = 5 per group). Bar represents the mean ± SEM. Multiple cell death pathways were upregulated in the brains of POWV-infected mice. Genes associated with death receptor, necroptosis and pyroptosis were highly enriched in POWV infected mouse brain at 6 and 8 dpi (Figure 5D). Genes involved in death receptor signaling such as Daxx, Fas, Parp3, Parp9, Parp12, and Parp14 were highly increased in infected brains. Genes involved in the necroptosis and pyroptosis pathways were significantly enriched in infected mice brains. These genes include Gsdmd, Casp1, Casp4, Casp8, Gbp4, Gbp5, Gbp7, Tlr2, and Tlr9. In general, genes involved in neuroinflammation, and cell death pathways were minimally induced at 3 dpi, increased significantly at 6 and 8 dpi, followed by a notable decrease by 21 dpi. Notably, the activation/inhibition patterns of neuroinflammation and cell death pathways coincided with the detected infectious viral titers in the brain. Validation of differential gene expression by qRT-PCR and multiplex cytokine/chemokine profiling To validate RNA-Seq data, we performed RT-qPCR on selected genes that were differentially expressed in POWV-infected brains. Consistent with RNA seq data, there was a significant upregulation of Ccl2, Il6, Casp7, Casp8, Casp4, and Gsdmd genes in POWV-infected brains (Figure 5E). Therefore, RNA-seq data accurately infer the biological changes in host gene expression induced by POWV infection. Next, we examined the dysregulation of cytokines and chemokines at the protein level following POWV infection through a multiplex immunoassay. Enhanced gene expression directly correlated with upregulated protein expression of multiple proinflammatory cytokines and chemokines. The upregulation of most proinflammatory mediators at 3 dpi was minimal except for G-CSF, IL6, and CXCL10. In contrast, at 6 dpi, we observed a significant upregulation in multiple cytokines and chemokines including IFNγ, IL-1α, IL-6, IL12, CCL2, CCL3, CCL11, CXCL2, CXCL9, and CXCL10 which was sustained through 8 dpi (Figure 6). Cytokine and chemokine profiling in POWV-infected brain. Brains were harvested from mock- and POWV-infected mice at days 3, 6, and 8 post infection. Cytokine and chemokine protein levels were measure by a multiplex immunoassay. Each data point represents an individual mouse (n = 6 mice per group). Middle horizontal bar indicates the mean and value error bars are SEM. P-values were calculated by One-way ANOVA followed by Dunnett’s multiple comparisons test. (*p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001). POWV infection induces glial activation, leukocyte infiltration, and cell death in the brain To further characterize POWV-induced immunopathology and validate the RNA-seq findings, we performed immunofluorescence staining of brain tissues collected from infected mice. Immunofluorescence analysis demonstrated marked activation of microglia in POWV-infected brains, as evidenced by increased IBA1-positive staining compared to mock-infected controls (Figure 7A). Infected brain sections showed increased IBA1 expression in the brain cortex and cerebellum of POWV-infected mice. In addition, astrocyte activation was evaluated by GFAP staining. POWV-infected brain tissues exhibited significantly increased GFAP immunoreactivity relative to mock-infected controls, indicating robust astrocyte activation and reactive gliosis following infection (Figure 7B). To determine whether immune cell infiltration occurred in the infected brain, sections were stained for CD45. POWV-infected tissues displayed substantially elevated CD45-positive staining compared to mock controls, indicating infiltration of peripheral leukocytes into the CNS (Figure 7C). Immunofluorescence analysis of POWV-infected brain tissues showing glial activation, and leukocyte infiltration. (A) Mock- and POWV-infected brain sections stained with IBA1 (red) and DAPI (blue). Scale bars are 150μm and 75μm on the original and enlarged images, respectively. (B) Immunofluorescence imaging of brain sections stained with GFAP (red) and DAPI (blue). Scale bars are 250μm and 75μm on the original and enlarged images, respectively. Representative images are shown (n=4 per group). (C) Immunofluorescence imaging of POWV-infected brain sections stained with Neurofilament-L (green), CD45 (red) and DAPI (blue). Representative images are shown (n=4 per group). We next evaluated POWV-induced cell death using TUNEL staining of brain tissues. POWV-infected mice showed markedly increased numbers of TUNEL-positive cells throughout the brain compared to mock-infected controls, demonstrating extensive virus-associated cellular injury and apoptosis (Figure 8A). Consistent with these observations, immunofluorescence staining revealed elevated caspase-3 expression in infected brain tissues, indicating activation of caspase-mediated apoptotic pathways during POWV infection (Figure 8B). In addition to apoptosis, POWV infection also induced increased p16 expression in brain tissues compared to mock controls (Figure 8C), suggesting activation of senescence-associated pathways and potential long-term cellular dysfunction following infection. Together, these results demonstrate that POWV infection induces significant neuroinflammation characterized by activation of microglia and astrocytes, infiltration of peripheral immune cells, and extensive apoptotic and senescence-associated cellular damage within the CNS. Apoptosis and cellular senescence following POWV infection in the brain. (A) TUNEL assay (green) was conducted on brain sections from mock- and POWV-infected mice to detect apoptotic cells. (B) Mock- and POWV-infected brain sections stained with cleaved-Caspase-3 (green) and DAPI (blue). C) p16 (green) staining was performed on brain sections from mock- and POWV-infected mice to assess cellular senescence. Representative images are shown (n=4 per group). Discussion Cases of POWV in the North America have been steadily increasing for the past 20 years. In this study, we examined the pathogenesis of POWV infection in C57BL/6J mice. Our study directly compared multiple routes of infection and its relation to morbidity and mortality. We demonstrated that a peripheral POWV infection by all routes leads to severe neuroinvasive disease in mice. Our data also demonstrated a dose-dependent mortality in POWV-infected mice through SC infection. We showed that POWV effectively replicates in C57BL/6 mice, where replication and dissemination resulted in peripheral and neurotropic phases. Importantly, mice that survived the acute POWV encephalitis still had detectable virus at 21 dpi. Additionally, we found several dysregulated genes in the brain during the acute and late stages of disease, including genes associated with interferon signaling, inflammation and cell death signaling. Notably, mice infected with POWV that survived the acute encephalitis phase showed persistent changes in the brain. POWV infection in humans can result in severe encephalitis, cognitive impairment and paralysis, with about 50% of patients reporting neurological sequalae (24–26). Previous reports have shown that inbred laboratory mouse strains, such as BALB/c and C57BL/6, develop neurological disease when peripherally infected with POWV (16–18, 27–29). Our study is the first to directly compare multiple routes of infection. We showed that IC, IP, and IV infection routes resulted in more rapid and severe disease compared to SC footpad infection. Our data clearly demonstrated a dose-dependent response in POWV-infected mice through SC infection. A time-course study showed rapid development of viremia beginning at 2 dpi followed by neuroinvasion and robust virus replication in the brain. Consistent with other reports (16, 17, 30, 31), our study also demonstrated significant brain pathology characterized by perivascular cuffing, glial cell activation and neuronal degeneration. Additionally, all surviving mice in this study had detectable levels of infectious virus and viral RNA in the brain at 21 dpi. These surviving mice also displayed mild neurological signs such as limb weakness or walking abnormalities. These findings indicate persistence of infectious POWV in the brains of surviving mice. Previous reports have showed viral persistence in the brain of mice infected with closely related deer tick virus (DTV) (32). Similarly, persistence of infectious WNV was observed in mice with subclinical infection (33). Previous reports have shown that during encephalitic viral infections, the loss of integrity of the BBB may contribute to virus access to the brain or may be the result of pathogenic viral replication in the BBB (34, 35). Our findings demonstrate that POWV infection results in significant disruption of BBB integrity, which likely contributes to viral neuroinvasion and CNS disease. Several neurotropic flaviviruses, including WNV, JEV, and TBEV, have previously been shown to alter BBB permeability during infection (12, 36–38). We previously reported that WNV can initially enter the CNS without major BBB disruption, while subsequent viral replication and neuroinflammatory responses lead to progressive BBB damage during later stages of disease (37, 38). Similar findings have been reported for TBEV and JEV, where BBB disruption occurs during later stages of infection when high viral titers are already present in the brain (12, 39). In contrast, the current study demonstrates that POWV infection induces early BBB dysfunction following peripheral infection, coinciding with systemic viremia. Overall, we demonstrated that a peripheral POWV infection induced disruption of the BBB in mice that occurred early during infection and correlated with virus entry to the CNS. To our knowledge, this is the first study to experimentally demonstrate BBB disruption during POWV infection. Whole brain transcriptomics provided additional evidence of POWV-induced acute and persistent transcriptional changes in the brain. We observed a dramatic dysregulation in genes involved in the immune response, primarily driven towards interferon-dependent response. Our results suggested that POWV induces a robust IFN response that remained high in all the surviving mice. Similar interferon-mediated responses have been reported in mice infected with other flaviviruses, including TBEV, LGTV, WNV, and JEV (22, 40–43). Type I IFNs (IFNαs and IFNβ) and their role during flavivirus infection have been well studied. In models with a deficient or compromised IFN response, neurons and astrocytes were more susceptible to infection with various flaviviruses. In rodent models, IFN response limits the spread of viral CNS infection and protects mice from lethal encephalitis by WNV, LGTV, and TBEV infection (40, 44). Similarly, in our recent studies, mice deficient in IFN signaling exhibited increased mortality, enhanced viral tropism, and heightened systemic inflammation after POWV infection compared with wild-type mice (41). Several inflammatory mediators are widely implicated in the pathogenesis of neurological disease (44–48). Consistently, we detected increased levels of several cytokines and chemokines that coincided with increased virus titers in the CNS and BBB disruption at 6 and 8 dpi. Notably, inflammation was observed until 21 dpi in the brain, albeit at lower levels. For example, we detected high expression of Ccl2 and Ccl5 at 6 and 8 dpi that persisted until 21 dpi. Our multiplex assay confirmed the release of these inflammatory mediators at the protein levels. These inflammatory molecules may further activate downstream cell death signaling pathway(s) in neurons ultimately resulting in neuronal death and/or activate glial cells, which can further exacerbate neuroinflammation (42, 44, 45, 47). Previous studies with POWV, TBEV, and LGTV have demonstrated that tick-borne flaviviruses induce robust neuroinflammatory responses characterized by microglial activation, leukocyte infiltration, and neuronal injury (12, 17, 40, 48). Like these reports, our study demonstrates that POWV infection induces activation of microglia and astrocytes, along with infiltration of peripheral immune cells into the CNS. Notably, we observed increased IBA1 and GFAP expression together with infiltration of CD45-positive leukocytes in POWV-infected brains. GFAP overexpression is linked to astrogliosis and has been observed in spinal cord cultures infected with WNV as well as autopsied neural tissues from humans with WNV encephalomyelitis (30, 31). We also observed activation of multiple cell death pathways during the acute encephalitic phase. TUNEL staining revealed extensive neuronal apoptosis in infected brains, supported by increased caspase-3 expression, indicating activation of apoptotic pathways. In addition, POWV infection induced elevated p16 expression, suggesting activation of senescence-associated pathways and potential long-term cellular dysfunction. Together, these findings demonstrate that excessive neuroinflammation, apoptosis, and cellular senescence contribute to POWV-induced neuropathology and highlight the severe immunopathological consequences of POWV infection in the brain. In summary, our results demonstrated that a peripheral POWV infection by all routes leads to severe neuroinvasive disease in C57BL/6J mice. We showed that POWV infection results in rapid development of viremia, followed by neuroinvasion and viral persistence in the brain. Our data further demonstrated that POWV penetrates the brain parenchyma early after infection with concurrent disruption of the BBB permeability. In addition, brain transcriptomics provided additional information on gene dysregulation during the acute and late stages of a POWV infection in mice. Materials and methods POWV infection in mice In this study, POWV, LB strain (BEI resources, NR-51181), was used for all the experiments. This is a prototype POWV strain and has been used in multiple POWV pathogenesis studies. The complete genome of POWV, LB has been sequenced (GenBank: L06436). Virus was propagated once in BHK-21 cells. We used approximately equal numbers of male and female mice for our experiments. Six-week-old C57BL/6 mice were inoculated with POWV or PBS (mock) through various routes of infection. Subcutaneous infection was given through the rear footpad injecting 20 µl of virus or PBS. Intraperitoneal infection was administered through a shallow injection in the peritoneum and 100 µl was administered of virus or PBS to each animal. Intravenous infection was administered in the tail vein following dilation with a heat lamp. Each mouse received 100 µl of virus or PBS. For intracranial, mice were injected in center of the frontal lobe just above the center of the eyes 5 mm deep into the brain and administered 20 µl of virus or PBS. For the dose-response study, six-week-old C57BL/6 mice were inoculated with 100, 1,000, and 10,000 PFU subcutaneously via footpad and daily observations were documented for each group. In a separate experiment, mice were inoculated with 1,000 PFU of POWV or PBS (mock) subcutaneously and euthanized at various time points to measure the viral titers, histopathology and RNA sequencing. Upon time-point or clinical end point, animals were anesthetized with isoflurane and perfused with cold PBS. Brain tissues were collected, and flash-frozen in 2-methylbutane (Sigma, St. Louis, MO, USA) for RNA and protein analysis as described below. Virus titers in serum and brain homogenates were measured by plaque assay using Vero cells as described previously (41). Evans blue assay BBB integrity was evaluated by using the Evans blue dye exclusion test as described previously (49). Six-week-old C57BL/6 mice were inoculated with 1,000 PFU of POWV or PBS (mock) via the footpad route. On days 2, 3, and 6 after infection, control and infected mice were injected in the peritoneal cavity with 1 ml Evans blue dye [1% (w/v) in PBS]. One hour later, mice were anesthetized with isoflurane and then perfused with 10 ml of cold PBS. Brains were harvested and photographed. RT-qPCR After Tissue homogenization and lysis in RLT buffer, total RNA was extracted using a Qiagen RNeasy Mini kit (Qiagen, Germantown, MD, USA) following the manufacturer’s instructions. cDNA was synthesized from 1,000 ng of RNA using an iScriptTM cDNA synthesis kit (Biorad, Catalog # 1708891) (50). The cDNA was used for qPCR using SsoAdvanced™ Universal SYBR® Green Supermix (Biorad, Catalog # 1725271) to calculate gene fold change (Table 1) (51, 52). Fold change of the RNA copy number of genes for infected brains were calculated against control samples after normalizing to the Gapdh gene. Viral RNA levels were measured with primers and probes specific for POWV. Viral genome copies were calculated using a standard curve and expressed per µg of total RNA. Table 1Primer sequences used for RT-qPCR.PrimerForward sequence (5’->3’)Reverse sequence (5’->3’)Il6 (NM_000600)CCAGGAGCCCAGCTATGAACCCCAGGGAGAAGGCAACTGCcl2 (NM_011333)TCACCTGCTGCTACTCATTCACCATACAGCTTCTTTGGGACACCTGCTCasp7 (NM_007611)CGTGGGAACGATGACCGAGGATTCATCAACGAACGGCGCasp8 (NM_009812)GGCCACAGGTTACAGCTCTTGGCCACAGGTTACAGCTCTTCasp4 (NM_007609)CCTGCAGAGGTGGGAACTCAAGGTTGCCCGATCAATGGTGsdmd (NM_026960)ATCAAGGAGGTAAGCGGCAGCCTTCTCCCATGCCTGACAAPOWVGGCTGCAAATGAGACCAATTCCAGCGACACATCTCCATAGTC Primer sequences used for RT-qPCR. Immunohistochemistry Sagittal sections (5-μm) were cut from OCT- or Paraffin-embedded hemibrain tissues. Tissue sections were stained with hematoxylin and eosin (H&E) for histopathological evaluation (Abcam, Catalog # ab245880). Additionally, tissue sections were stained with dsRNA Antibody (J2), Alexa Fluor™ 488 (Fisher scientific, Catalog # NB272461), Anti-GFAP-Alexa Fluor 594 Anti-mouse (Cell Signaling, Catalog # 8152), NeuN mAb with Alexa Fluor® 555 Conjugate (Cell Signaling, Catalog # 23357), Anti-AIF-1/Iba1 Alexa Fluor 594 Anti-mouse (Novus biologicals, Catalog # D169192), Anti-cleaved Caspase-3, Alexa Fluor® 488 Conjugate (R&D Systems, Catalog # IC835G), CDKN2A/P16-INK4a Rabbit Polyclonal (BIOSSUSA, Catalog # BD06066741), Neurofilament-L (C28E10) Rabbit Monoclonal Antibody Alexa Fluor® 488 Conjugate (Cell Signaling, Catalog # 8024S) overnight at 4˚C. Terminal deoxynucleotidyl transferase (dUTP) nick end labeling (TUNEL) staining was performed on lung sections using an in-situ cell death detection kit (Roche, Catalog # 11684795910) following the manufacturer’s instructions. Images were acquired using the Invitrogen™ EVOSTM M5000 Cell Imaging System (Thermo Fisher Scientific, Norcross, GA, USA) (53–55). RNA-sequencing and ingenuity pathway analysis (IPA) Transcript quantification using RNA-Seq technology was performed for each infected mouse sample collected at 3, 6, 8, and 21 dpi and mock-infected controls (n=4/group) using Illumina platform (150 bp paired-end). After the read mapping against the mouse reference genome UCSC mm10, Stringtie was used for transcript assembly. Expression profiles were calculated for each sample and are presented as read count and normalization value which is based on transcript length and depth of coverage. The FPKM (Fragment per Kilobase of transcript per Million mapped reads) is used as a normalization value. The generated FPKM average value of each time point was compared to the average of the mock-infected controls to determine the number of significantly differentially expressed genes (DEGs). Poor-quality sequences were trimmed, and differential gene expression analysis was performed for each group using the cutoff values of Log2 fold change > |± 1.5| and a false discovery rate (FDR) or corrected p value (q value) of < 0.05. QIAGEN’s Ingenuity Pathway Analysis (IPA, QIAGEN Redwood City) was utilized to investigate the significantly DEGs at the level of biochemical pathways, molecular functions, diseases and biofunctions (51, 56). DEGs data sets were uploaded to IPA to perform functional analysis. Comparative analysis, and individual analysis were analyzed using IPA Knowledge Base which is the largest database of manually curated and experimentally validated physical, transcriptional, and enzymatic molecular interactions. Luminex assay Brain harvested from mice at days 3, 6, and 8 post infection, and mock-infected were homogenized in 1X PBS with protease inhibitor in the bullet blender (Next Advanced). Homogenates were tested using the MILLIPLEX MAP Mouse Cytokine/Chemokine Magnetic Bead Panel - Premixed 32 Plex - Immunology Multiplex Assay (Catalog # MCYTMAG-70K-PX32) as per manufacturer instructions as previously described (56–58). Statistical analysis Statistical analysis was performed using Prism software (GraphPad, version 10, San Diego, USA), with p<0.05 considered significant. Tests, number of animals and p values are indicated in each figure legend. Difference in survival was assessed through a Mantle-Cox test. We used one-way analysis of variance (ANOVA) followed by Dunnett’s multiple comparisons test or the non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparisons test to compare multiple groups. Genes with FDR value <0.05 and Log2FC > |±1.5| were considered significantly differentially expressed (DEGs).