RIPK3 promotes brain region-specific interferon signaling and restriction of tick-borne flavivirus infection

Innate immune signaling in the central nervous system (CNS) exhibits many remarkable specializations that vary across cell types and CNS regions. In the setting of neuroinvasive flavivirus infection, neurons employ the immunologic kinase receptor-interacting kinase 3 (RIPK3) to promote an antiviral transcriptional program, independently of the traditional function of this enzyme in promoting necroptotic cell death. However, while recent work has established roles for neuronal RIPK3 signaling in controlling mosquito-borne flavivirus infections, including West Nile virus and Zika virus, functions for RIPK3 signaling in the CNS during tick-borne flavivirus infection have not yet been explored. Here, we use a model of Langat virus (LGTV) encephalitis to show that RIPK3 signaling is specifically required in neurons of the cerebellum to control LGTV replication and restrict disease pathogenesis. This effect did not require the necroptotic executioner molecule mixed lineage kinase domain like protein (MLKL), a finding similar to previous observations in models of mosquito-borne flavivirus infection. However, control of LGTV infection required a unique, region-specific dependence on RIPK3 to promote expression of key antiviral interferon-stimulated genes (ISG) in the cerebellum. This RIPK3-mediated potentiation of ISG expression was associated with robust cell-intrinsic restriction of LGTV replication in cerebellar granule cell neurons. These findings further illuminate the complex roles of RIPK3 signaling in the coordination of neuroimmune responses to viral infection, as well as provide new insight into the mechanisms of region-specific innate immune signaling in the CNS.


Introduction
Flaviviruses are a family of positive sense RNA viruses which include several notable pathogens associated with neuroinvasive infection in humans, including West Nile virus (WNV), Zika virus (ZIKV), and Japanese Encephalitis virus [1].While nearly all major flaviviruses are transmitted by mosquito vectors, a small but significant number of flaviviruses are transmitted by ticks, including Tick-borne encephalitis virus (TBEV) and its close relatives that together make up a single TBEV serocomplex.Tick borne encephalitis is a significant and growing threat to public health, particularly in Europe and northern Asia, where TBEV constitutes the most prevalent tick-borne zoonotic disease [2][3][4].Notably, some TBEV strains elicit mortality rates up to 40% in humans [5], underscoring the urgent need to better understand the mechanisms underlying the pathogenesis of tick-borne flavivirus infections.
Effective control of flavivirus infection in the central nervous system (CNS) requires robust innate immune signaling in neural cells, particularly neurons, which are the predominantly infected cell type in most cases of flavivirus encephalitis [6][7][8][9].Effective type I interferon (IFN) signaling is of particular importance for innate control of viral replication in neurons [10][11][12].Notably, differences in type I IFN signaling across neural cell types and brain regions are associated with differential susceptibility to flavivirus infection.For example, previous reports suggest that the enhanced type I IFN signaling observed in hindbrain regions compared to the forebrain is an underlying determinant of the enhanced susceptibility of forebrain regions to WNV infection [12,13].However, the unique signaling mechanisms that promote differential IFN-mediated control of viral infection in the hindbrain have not been extensively characterized.
A potential regulator of neuronal IFN signaling during flavivirus infection is receptor interacting protein kinase-3 (RIPK3).RIPK3 is an enzyme traditionally associated with necroptosis, a form of lytic programmed cell death [14].Necroptosis occurs via the RIPK3-dependent activation of mixed lineage kinase domain like protein (MLKL), which forms oligomeric pore complexes that induce cellular lysis [15].However, many recent studies have identified complex roles for RIPK3 signaling in the coordination of inflammation, including the regulation of inflammatory transcriptional responses that occur independently of necroptosis [16][17][18][19][20][21][22][23][24].We and others have demonstrated that RIPK3 signaling in neurons is of particular importance for the control of neurotropic viral infections, as neuronal RIPK3 promotes a robust antimicrobial transcriptional program, including many IFN stimulated genes (ISGs), that restricts viral infection without inducing neuronal necroptosis [16,17].Other recent studies have identified unexpected roles for RIPK3 in the regulation of type I IFN signaling, via mechanisms which include the regulation of pattern recognition receptor signaling and protein kinase-R (PKR)mediated stabilization of Ifnb mRNA [18,19].
In this study, we interrogated roles for RIPK3 in controlling tick-borne flavivirus infection.To do so, we used Langat virus (LGTV), a naturally attenuated member of the TBEV serocomplex that can be studied under BSL2 containment.LGTV infection is neuroinvasive in rodents and results in mild clinical disease [25,26].Ripk3 -/-mice exhibited enhanced neurologic disease following subcutaneous LGTV infection, while Mlkl -/-mice were indistinguishable from littermate controls, suggesting a necroptosis-independent function for RIPK3 in restricting LGTV pathogenesis.Notably, Ripk3 -/-mice exhibited increased viral burden in the cerebellum, along with diminished expression of inflammatory chemokines and ISGs in the cerebellum, but not the cerebral cortex.In vitro analysis of cultured primary cortical and cerebellar cell types showed that pharmacologic inhibition of RIPK3 resulted in enhanced LGTV replication in cerebellar granule cell neurons but not in cortical neurons or in astrocytes and microglia derived from either brain region.Transcriptional profiling showed that RIPK3 signaling was uniquely required for the full induction of ISG expression in cerebellar granule cell neurons, demonstrating a previously unknown, region-specific function for RIPK3 in coordinating innate antiviral immunity within the CNS.

RIPK3 controls LGTV pathogenesis independently of MLKL and peripheral immunity
To assess the role of RIPK3 in controlling LGTV pathogenesis, we subcutaneously infected Ripk3 -/-mice, along with heterozygous littermate controls, with 3x10 4 plaque forming units (pfu) of the Malaysian LGTV strain TP21.We note that Ripk3 +/-animals do not exhibit haploinsufficiency and are routinely used as littermate controls in studies of this pathway [27][28][29].Control animals exhibited limited mortality following LGTV infection (Fig 1A ), consistent with previous reports [25,26].However, mice lacking Ripk3 expression exhibited a significantly accelerated and enhanced rate of mortality (Fig 1A).In addition, a higher proportion of Ripk3 -/-mice exhibited clinical signs of neurologic disease, including paresis or full hindlimb paralysis, by 14 days post infection (dpi) (Fig 1B), and this difference persisted to at least 21 dpi.These data suggest that Ripk3 is essential for restricting neuropathogenesis during LGTV infection.
To better understand this phenotype, we first assessed whether RIPK3 was required for early control of systemic infection.Spleens of infected mice exhibited low levels of LGTV RNA that were not impacted by Ripk3 expression (Fig 1C).To test whether Ripk3 -/-mice exhibited any deficiencies in peripheral immune responses, we performed flow cytometric analysis of major immune cell subsets in the spleens of infected animals at 8 dpi.Ripk3 -/-animals exhibited similar frequencies (Fig 1D .Both CD11c + MHCII + and F4/80 + antigen presenting cell subsets also exhibited similar rates of expression of the costimulation signal CD80 between genotypes (Fig 1O and 1P).These data suggest that Ripk3 -/-mice mounted normal peripheral immune responses to subcutaneous LGTV challenge, similar to our previous observations with WNV and ZIKV [16,17].Thus, the increased pathogenesis observed in mice lacking Ripk3 was unlikely to arise from a failure in peripheral virologic control.
A potential mechanism by which RIPK3 signaling might restrict LGTV pathogenesis is through the induction of necroptosis in infected cells.We thus tested whether loss of the necroptotic executioner molecule MLKL would impact disease course following subcutaneous LGTV infection.Notably, Mlkl -/-mice exhibited no difference in either survival or development of clinical disease signs compared to littermate controls (Fig 2A and 2B).We saw similarly that Mlkl -/-did not exhibit altered splenic viral burden at 8dpi (Fig 2C).Flow cytometric analysis also revealed essentially identical numbers and frequencies of all major immune cell subsets in the spleen at this time point (Fig 2C -2P).Multistep growth curve analysis also demonstrated that neither RIPK3 nor MLKL impacted the low levels of LGTV replication observed in primary leukocyte cultures, including bone marrow derived macrophages and dendritic cells (S1A and S1B Fig).These data suggest that MLKL, and therefore necroptosis, is not a major contributor to peripheral virologic control or overall disease pathogenesis in the setting of LGTV infection, and thus that RIPK3 exerts its protective effect in this model through an alternative mechanism.

RIPK3 is required for CNS-intrinsic restriction of LGTV infection
Because we did not observe differences in peripheral virologic control in Ripk3 -/-mice, we next questioned whether RIPK3 acted in a CNS-intrinsic manner to limit LGTV infection.To

RIPK3 promotes neuronal chemokine expression in a region-specific manner following LGTV infection
We and others previously showed that neuronal RIPK3 signaling was required for the expression of key inflammatory chemokines that served to restrict WNV pathogenesis by coordinating the recruitment of leukocytes into the infected CNS.We thus questioned whether RIPK3 also promotes chemokine expression in the CNS during LGTV infection.Surprisingly, transcriptional profiling in the cerebral cortex of Ripk3 -/-mice following subcutaneous LGTV infection revealed no differences in expression of major chemokines compared to littermate controls (Fig 4A).However, we did observe significantly diminished chemokine responses in cerebellar tissues derived from Ripk3 -/-animals (Fig 4B).We confirmed differential expression of CXCL10 in the cerebella but not cerebral cortices of Ripk3 -/-mice at the protein level by

ELISA (S2A and S2B Fig).
To understand which cell types were driving this region-specific deficit in chemokine expression, we next cultured primary neurons and astrocytes derived specifically from either cerebral cortex or cerebellum and infected with LGTV, with or without a small molecule inhibitor of RIPK3 (GSK 872).Consistent with our in vivo findings, blockade of RIPK3 in cerebral cortical neurons did not impact chemokine expression following LGTV infection (Fig 4C).In contrast, infected cerebellar granule cell neuron cultures exhibited significantly diminished chemokine expression when RIPK3 was inhibited by GSK 872 (Figs 4D and S2C).Notably, we did not observe a dependence on RIPK3 for the expression of chemokines in astrocytes derived from either region (Fig 4E and 4F).These data suggest that RIPK3 serves an unexpected, region-specific transcriptional function in neurons of the cerebellum during neuroinvasive LGTV infection.

RIPK3 is not required for immune cell recruitment to the LGTV-infected CNS
We next questioned whether diminished chemokine expression in the cerebellum of Ripk3 -/- mice would result in a failure to recruit antiviral leukocytes into this brain region.We thus performed flow cytometric analysis of leukocytes derived from either cerebral cortex or cerebellum following subcutaneous LGTV infection.Remarkably, we saw no evidence of changes in lymphocyte recruitment in either brain region of Ripk3 -/-mice compared to littermate controls on either 6 or 8 dpi (Fig 5A).This lack of difference extended across all major CD45 hi infiltrating leukocyte subsets, including CD4 + and CD8 + T cells ( .We similarly did not observe differences in numbers of CD45 lo microglia (Fig 5I ), suggesting no major differences in microglial proliferation between genotypes in either region.These data suggested that, despite significant differences in the expression of major leukocyte chemoattractants in the cerebellum, differences in immune cell recruitment did not account for the increased pathogenesis observed in Ripk3 -/-mice during LGTV infection.

RIPK3 promotes cell-intrinsic restriction of LGTV replication in cerebellar neurons
Given these observations, we next questioned whether Ripk3 -/-mice fail to control LGTV infection due to impaired innate immune restriction of LGTV replication.Assessment of viral burdens in brains of Ripk3 -/-mice following subcutaneous LGTV infection revealed that Ripk3 -/-mice exhibited significantly elevated CNS viral titers, particularly in the cerebellum, at both 8 and 12 dpi (Fig 6A).In contrast, Mlkl -/-exhibited no such difference in viral burden in either brain region (Fig 6B).Differences in viral burden did not appear to be linked to deficits in blood-brain barrier integrity, as both Ripk3 -/-mice and littermate controls exhibited similar levels of sodium fluorescein extravasation into the CNS following infection (Fig 6C).We thus questioned whether RIPK3 was required for cell-intrinsic restriction of viral replication in susceptible CNS cell types.While LGTV preferentially infects neurons in vivo [25], we nevertheless tested a panel of primary CNS cell types to assess whether RIPK3 may also influence LGTV tropism.Multistep growth curve analysis in primary CNS cells revealed that pharmacologic inhibition of RIPK3 had no effect on LGTV replication in neurons derived from cerebral cortex (Fig 6D).In contrast, inhibition of RIPK3 significantly enhanced LGTV replication in primary cerebellar granule cell neurons cultures (Fig 6E).This effect was unique to neurons, as GSK 872 treatment had no impact on LGTV replication in primary astrocytes derived from either brain region (Fig 6F -6G).LGTV replication was not detectable by plaque assay in primary microglial cultures; however, we did not observe Ripk3-dependent differences in LGTV RNA in primary microglia derived from either brain region (S3A and S3B Fig) .Together, these data suggested that the enhanced pathogenesis observed in Ripk3 -/-mice was due to a specific failure to control infection in neurons of the cerebellum, resulting in enhanced overall CNS viral burden.

RIPK3 potentiates Type I IFN signaling in cerebellar neurons during LGTV infection
Our previous observation of diminished chemokine expression in cerebellar neurons derived from Ripk3 -/-mice suggested that these cells may exhibit broader deficits in innate immune signaling, resulting in poor control of LGTV replication.Previous work has shown that cerebellar granule cell neurons exhibit enhanced type I IFN signaling compared to cortical neurons [12].We, therefore, next questioned whether IFN signaling was perturbed in the cerebellum of mice lacking RIPK3 expression.Transcriptional profiling in brain tissues following subcutaneous LGTV infection revealed that, indeed, the cerebella of Ripk3 -/-mice exhibited diminished expression of many ISGs known to be critical for control of flavivirus replication [30][31][32][33][34][35], including Ifit1, Isg15, Mx1, Mx2, Oas1b, and Rsad2, while this phenotype was not observed in the cerebral cortex (Fig 7A and 7B).Similar analyses in primary cell cultures confirmed that cerebellar granule cell neurons, but not neurons derived from cerebral cortex, exhibited diminished expression of ISGs when RIPK3 signaling was blocked via GSK 872 treatment (Fig 7C and 7D).In contrast, we observed little to no impact of RIPK3 blockade on ISG expression in astrocytes derived from either brain region (Fig 7E and 7F).Surprisingly, we did observe that lack of Ripk3 expression diminished ISG expression in primary microglia, although this phenotype occurred in both cerebral cortical and cerebellar microglia cultures (S3B and S3C Fig) , suggesting that variations in microglial ISG expression were not the source of regional differences in virologic control in the brain.Differences in ISG expression between regional neuron types were not driven by differential rates of cell death, as pharmacologic blockade of RIPK3 did not impact viability of either neuron type following LGTV infection (S4 Fig) .Together, these data demonstrate that RIPK3 signaling is required for the robust induction of type I IFN responses in neurons of the cerebellum, which is required for cell-intrinsic restriction of LGTV replication.
To better understand the role of RIPK3 signaling in potentiating ISG expression, we next questioned whether RIPK3 acts downstream of IFN receptor signaling.To assess this, we treated neuron cultures with exogenous IFNβ for 1 hour following pretreatment with GSK 872 or vehicle control.As expected, IFNβ treatment resulted in robust induction of multiple ISGs (Figs 8A and S5A).However, pharmacologic blockade of RIPK3 did not impact ISG expression induced by IFNβ treatment in either cerebellar granule cell neurons (Fig 8A ) or in cerebral cortical neurons (S5A Fig) , suggesting that RIPK3 likely does not act directly downstream of the type I IFN receptor (IFNAR) to modulate gene expression and/or that type I IFN alone is not sufficient to induce RIPK3 activation.We next tested the alternative hypothesis that RIPK3 regulates IFN signaling during LGTV infection by directly influencing the expression of IFN ligands and receptors.Surprisingly, transcriptional analysis revealed that pharmacologic blockade of RIPK3 did not influence the expression of the type I IFN ligands Ifna and Ifnb in cerebellar granule cell neurons following infection (Fig 8B).In contrast, GSK 872 treatment significantly blunted infection-induced upregulation of IFN receptor subunits, including the type I IFN receptor subunits Ifnar1 and Ifnar2, and the type II IFN receptor subunits Ifngr1, and Infgr2.Importantly, we did not observe this RIPK3-dependency in IFN receptor expression in cerebral cortical neuron cultures (S5B Fig) , suggesting that RIPK3 functions uniquely in cerebellar granule cell neurons to enhance type I IFN signaling during LGTV infection.
To further investigate a role for RIPK3 in IFN-mediated gene expression in cerebellar neurons, we next performed experiments in which we blocked RIPK3 activity with or without simultaneous blockade of type I IFN signaling using a neutralizing antibody against IFNAR1.We reasoned that this paradigm would allow us to assess the differential influence of RIPK3 on IFNAR-dependent and IFNAR-independent gene expression following LGTV infection.Perhaps unsurprisingly, we observed that expression of most ISGs was completely dependent on IFNAR1 signaling, making it difficult to distinguish a specific role for RIPK3 in the absence of intact type I IFN signaling (S5C Fig) .We thus identified several alternative inflammatory genes whose expression was either completely (Cxcl1) or partially (Ccl2 and Il6) independent of IFNAR1 signaling following infection.Notably, pharmacologic blockade of RIPK3 only impacted the IFNAR1-depenent portion of the induced expression of these genes, while having no effect on the IFNAR1-independent portion, as indicated by a lack of effect in αIFNAR1-treated cultures (Fig 8C).Together, these data further support our observation of synergistic signaling between type I IFN and RIPK3 signaling in cerebellar granule cell neurons during LGTV infection.
To confirm these observations in vivo, we next performed immunohistochemical staining of IFNAR1 in cerebella of mice at 8 days following subcutaneous LGTV challenge.We observed significantly diminished colocalized signal of IFNAR1 on MAP2 expressing neurons within the granule cell layer of the cerebellum in Ripk3 -/-mice compared to littermate controls (Fig 9A and 9B), confirming that RIPK3 promotes interferon receptor expression in these cells at the protein level.To further confirm whether RIPK3 promoted ISG expression in these cells, we acutely isolated cerebellar granule cell neurons ex vivo (Fig 9C ) following intracranial LGTV infection.Isolated cells expressed high levels of the glutamate receptor genes Gria4 and Grid2, but low to undetectable levels of the glial cell-associated genes Gfap and Aif1, confirming our isolates were highly enriched for neurons (S6 Fig) .Gene expression analysis showed that ISG induction was significantly diminished in cerebellar neurons of Ripk3 -/-mice compared to controls (Fig 9D ), consistent with our previous findings using primary cultures.These data further support our finding that RIPK3 is required for robust induction of interferon signaling in cerebellar neurons during LGTV infection.

Discussion
Our findings identify a previously unknown function for RIPK3 in the coordination of brain region-specific innate immunity.The study of regional differences in neuroimmune signaling is a growing field, and there is accumulating evidence to suggest that resident neural cells exhibit differential responses to viral infection and cytokine stimulation across distinct anatomical regions of the CNS [36][37][38][39].Neurons and astrocytes in the cerebellum, in particular, have been shown to exhibit higher responsiveness to stimulation by type-I IFN, as well as to express higher basal levels of pathogen sensor molecules compared to other brain regions, suggesting a key evolutionary importance of innate antiviral defense in this tissue [12,13].This regional difference in type I IFN signaling appears to underlie, at least in part, the relatively lower susceptibility of the cerebellum to flavivirus infection compared to susceptible regions of the forebrain, such as the cerebral cortex and hippocampus.Our findings also underscore the particular importance of IFN signaling in cerebellar granule cell neurons, which account for over 99% of neurons in the cerebellum and more than 50% of the neurons in the entire brain [12,40].However, the molecular mechanisms that determine the enhanced innate immune signaling observed in the cerebellum remain poorly understood.Our study suggests that RIPK3 signaling is required for the robust induction of ISG expression in cerebellar neurons during LGTV infection, although ongoing work is needed to understand the specific signaling interactions that mediate this effect.
Previous studies have described a highly complex interplay between RIPK3 and type I IFN signaling that varies significantly by cell type and disease model [17][18][19]41].It is relatively clear that type I IFN signaling is capable of activating RIPK3 through various mechanisms, resulting in necroptosis and/or necroptosis-independent transcriptional activation [41][42][43][44].However, how RIPK3 operates upstream of (or synergistically with) type I IFN signaling to influence expression of ISGs is less clear.We and others have shown that ISG expression is significantly diminished in a variety of settings when RIPK3 signaling is ablated [17,18], including in cerebellar granule cell neurons during LGTV infection in this study.One possible explanation for this effect is RIPK3-mediated activation of NF-κB, a transcription factor strongly associated with RIPK signaling with known roles in potentiating type I IFN signaling and ISG expression [22,[45][46][47].We and others also previously showed that RIPK3 activation in cortical neurons following ZIKV infection leads to interferon regulatory factor 1 (IRF1) activation, which was required for expression of at least a subset of RIPK3-induced genes in that setting, although this effect is likely indirect, as IRF1 is not a known RIPK3 substrate [17].Additional work will be needed to fully characterize the regulatory mechanisms that are invoked in the interplay between RIPK3 and type I IFN signaling in the CNS.
Our study also further expands our understanding of the necroptosis-independent functions for RIPK3 signaling in the CNS.Many studies have now firmly established the importance of RIPK3 in promoting host defense through mechanisms independent of its canonical role in necroptosis [16][17][18][19][20][21].However, these necroptosis-independent functions appear to vary significantly by disease state, including CNS infection with distinct neuroinvasive flaviviruses [48,49].We and others previously showed that the primary role for RIPK3 in restricting WNV encephalitis was the induction of chemokine expression and the recruitment of antiviral leukocytes into the infected CNS [16].Notably, while we did observe RIPK3-mediated chemokine expression in the cerebellum during LGTV infection, this chemokine expression was apparently dispensable for CNS immune cell recruitment.These findings highlight the contextdependent nature by which RIPK3 seems to regulate neuroinflammation.Notably, the RIPK3-dependent neuronal transcriptional program includes both pro-and anti-inflammatory mediators [17,49].The centrality of IFN signaling in our current study is also notable, as we and others have previously shown that type I IFN suppresses leukocyte infiltration into the CNS during flavivirus encephalitis [13].Thus, diminished IFN signaling in LGTV-infected Ripk3 -/-mice may counterbalance the reduced expression of leukocyte chemoattractants in the overall regulation of CNS immune cell recruitment.
Unlike our previous findings using a model of WNV encephalitis [16], the RIPK3-dependent transcriptional activation of antiviral effector genes, including ISGs, was required for cell-intrinsic restriction of LGTV replication in neurons, a phenotype more similar to our findings with ZIKV [17], although we did not observe evidence for a regional specification of this response during ZIKV infection.In contrast to these observations, Bian and colleagues have observed quite distinct phenotypes in a model of JEV encephalitis, wherein both RIPK3 and MLKL appeared to exacerbate rather than restrict disease pathogenesis [50,51].RIPK3 also appeared to suppress rather than promote ISG expression in JEV infected neurons.At present, the factors that determine such distinct outcomes of RIPK3 signaling across this family of closely related viruses and among distinct CNS regions and cell types are mysterious.However, we speculate that the evolutionary pressures that have shaped differential innate immune signaling across distinct regional populations of neurons, such as enhanced pattern recognition receptor signaling in cerebellar granule cell neurons [12], have also likely driven convergent, context-specific functional specialization of Ripk3, a gene which exhibits high rates of evolution and strong positive selection across mammalian genera [52].However, further work is needed to fully define the molecular mechanisms that govern such specializations, as well as to assess whether regional differences in RIPK3 function in the CNS are conserved across species.

Ethics statement
All procedures involving animal care and use were approved by and conducted in accordance with the Rutgers University Institutional Animal Care and Use Committee (IACUC) under protocol number 201900016.

Mouse lines
Ripk3 -/- [53] and Mlkl -/- [54] mouse lines were bred and housed under specific-pathogen free conditions in Nelson Biological Laboratories at Rutgers University.Ripk3 -/-mice were generously provided by Genentech, Inc.Wild-type C57BL/6J mice were either obtained commercially (Jackson Laboratories) or bred in-house.Mice used for subcutaneous infections were 5 weeks old; mice used for intracranial infections were 8-15 weeks old.

Mouse infections and tissue harvesting
Isoflurane anesthesia was used for all procedures.Mice were inoculated subcutaneously (50uL) with 3x10 4 PFU or injected intracranially (10uL) with 50 PFU of LGTV-TP21 using insulin syringes (BD Medical, #BD-329461).At appropriate times post infection, mice underwent cardiac perfusions with 30 mL cold sterile 1X phosphate-buffered saline (PBS).Extracted tissues were weighed and homogenized using 1.0 mm diameter zirconia/silica beads (Biospec Products, #11079110z) in sterile PBS for plaque assay or TRI Reagent (Zymo, #R2050-1) for gene expression analysis.Homogenization was performed in an Omni Beadrupter Elite for 2 sequential cycles of 20 s at a speed of 4 m/s.
Macrophages and dendritic cells were isolated from bone marrow of euthanized mice.Femurs were isolated and bone marrow pushed out using a sterile needle and syringe loaded with RPMI supplemented with 10% FBS, 1% Penicillin-Streptomycin-Glutamine, 1% HEPES, 1% Glutamax (ThermoFisher, #35050061).Bone marrow was plated into non-cell-culture treated 10cm petri dishes in 8mL supplemented RPMI medium containing either 20ng/mL recombinant M-CSF (Peprotech, #315-02) or 20ng/mL recombinant GM-CSF (Peprotech, #315-03) and 20ng/mL IL-4 (Peprotech, #214-14) for differentiation into macrophages or dendritic cells, respectively.Cells were fed with additional medium containing the appropriate cytokines four days later and used for experiments at 6-7 DIV.Cells were seeded into cell-culture treated dishes prior to experimentation.For viral replication determination, all cultures were infected with LGTV TP21 at an MOI of 0.01.For qRT-PCR experiments, cortical and cerebellar neuron cultures were infected at an MOI of 0.5, while astrocyte cultures were infected using an MOI of 0.01.The pharmacologic inhibitor of RIPK3, GSK 872, was added to PLOS PATHOGENS Regional CNS RIPK3 signaling restricts LGTV infection cultures at 100nM for 2 hours prior to infection or subsequent treatments.Interferon-β was added to neuron cultures at 10ng/mL for 1 hour prior to harvesting cell lysates.IFNAR-1 monoclonal antibody (MAR1-5A3, Leinco Technologies) or isotype control (GIR-208, Leinco Technologies) were added to cultures at 5μg/mL 45 minutes prior to Langat virus infection.

Ex vivo cerebellar granule cell neuron isolation
Adult cerebellar granule cell neurons were isolated using a Percoll gradient as previously described [55], with minor modifications.Adult cerebella were dissociated using an Adult Brain Dissociation Kit (Miltenyi, #130-107-677).Cells were layered over a step gradient of 60% and 35% isotonic Percoll (Cytiva, #17-0891-02) and centrifuged at 2370g for 20 minutes.The middle fraction consisting of granule cells was carefully collected, and acutely isolated cells were immediately lysed for RNA extraction and were not cultured.

Quantitative real-time PCR
Total RNA from harvested tissues was extracted using Zymo Direct-zol RNA Miniprep kit, as per manufacturer instructions (Zymo, #R2051).Total RNA extraction from cultured cells, cDNA synthesis, and subsequent qRT-PCR were performed as previously described [22,56].Cycle threshold (CT) values for analyzed genes were normalized to CT values of the housekeeping gene 18 S (CT Target − CT 18S = ΔCT).Data from primary cell culture experiments were further normalized to baseline control values (ΔCT experimental − ΔCT control = ΔΔCT (DDCT)).A list of primers used in this study can be found in S1 Table.

In vivo assessment of blood brain barrier permeability
In vivo assessment of blood brain barrier permeability was carried out as described [57].Mice were injected intraperitoneally with 100uL of 100mg/mL fluorescein sodium salt (Sigma, #F6377) dissolved in sterile 1X PBS.After 45 minutes, blood was collected followed by cardiac perfusion.Tissues were dissected and homogenized in 1X PBS as described above.Serum and supernatant from homogenized tissues were incubated overnight at 4˚C with 2% Trichloroacetic acid solution (Sigma, #T0699) at a 1:1 dilution.Precipitated protein was pelleted by 10 minutes of centrifugation at 2,823xg at 4˚C.Supernatants were diluted with borate buffer, pH 11 (Sigma, #1094621000) to achieve a neutral pH.Fluorescein emission at 538nm was measured for samples in an optically clear black-walled 96-well plate (Corning, #3904) using a Spectra-Max iD3 plate reader (Molecular Devices, San Jose, CA).Tissue fluorescence values were standardized against plasma values for individual mice.

Statistical analysis
Normally distributed data were analyzed using appropriate parametric tests: two-way analysis of variance (ANOVA) with Sidak's correction for multiple comparisons and Log-rank (Mantel-Cox) test for survival comparison, both using GraphPad Prism Software v8 (GraphPad Software, San Diego, CA).Chi square tests for comparison of clinical disease signs was performed using Excel v2211 (Microsoft).P < 0.05 was considered statistically significant.
) and total numbers (Fig 1E and 1F) of CD4 and CD8 T cells among all splenocytes compared to littermate controls, as well as similar rates of CD44 expression (a key T cell activation marker) across both subsets (Fig 1G and 1H).Numbers of B cells (Fig 1I) and natural killer (NK) cells (Fig 1J) were also similar between genotypes.In the myeloid compartment, we observed similar numbers of CD11c + MHCII + dendritic cells (Fig 1K) between genotypes, as well as similar numbers of myeloid subsets expressing F4/80 (Fig 1L), Ly6G (Fig 1M), and Ly6C (Fig 1N)

Fig 9 .
Fig 9. RIPK3 promotes IFN signaling in cerebellar neurons in vivo.A-B) Representative images (A) and quantification (B) of IHC analysis of IFNAR1 expression on MAP2 expressing neurons in cerebella of mice at 8 dpi (subcutaneous).Colocalized signaling between IFNAR1 and MAP2 shown in yellow.Traces outline the granule cell layer of the cerebellum.Scale bar = 100um.Quantification represents intensity of colocalized signal normalized to the total size of the granule cell layer in each image.C) Experimental schematic depicting ex vivo isolation of adult cerebellar granule cell neurons (GCNs).Image made with BioRender.D) Transcriptional expression of indicated genes in isolated GCNs derived from mice of indicated genotypes at 4 dpi (intracranial).ns, not significant.*p<0.05,****p < 0.0001.Some figure elements were made with BioRender.https://doi.org/10.1371/journal.ppat.1011813.g009