Glenn F Rall, PhD
Office Phone: 215-728-3617
Lab Phone: 215-728-3677
Measles Virus Trafficking In, and Spread Among, Neurons of the NS
Central nervous system diseases that develop following acute measles virus (MV) infections, such as subacute sclerosing panencephalitis, are characterized by a lack of extracellular virus, despite abundant viral RNA. Similarly, using transgenic mice in which the MV receptor CD46 is targeted to CNS neurons, we have previously shown that MV challenge results in inter-neuronal spread with negligible extracellular virus. Furthermore, neuron-to-neuron spread of MV through the CNS in this mouse model occurs without syncytia formation, the typical mode of MV transmission in nonneuronal cells. Our work on the spread of MV in NSE-CD46 transgenic mice and in primary neurons explanted from these mice suggests that MV spreads trans-synaptically by a mechanism independent of the interaction of the hemagglutinin protein (H) with CD46. Studies addressing the role of the fusion protein (F) in mediating trans-synaptic spread showed that fusion inhibitory peptide (FIP) and Substance P inhibited MV trans-synaptic spread. Additional work revealed that CD46-expressing mice lacking the receptor for Substance P, neurokinin-1, as well as CD46-expressing mice treated with aprepitant, a pharmacological inhibitor of neurokinin-1, were resistant to MV infection in vivo. These studies underscored a role for the MV fusion protein in mediating trans-synaptic spread. Current research is centered on determining the downstream consequences of the interaction of MV-F with neurokinin-1. Preliminary data indicates that survival pathways, characterized by activated p42/p44 MAPK (ERK), are activated by MV following addition to primary hippocampal neurons, similar to that seen following addition of the natural neurokinin-1 ligand Substance P (SP) to primary hippocampal neurons. This research will continue to provide insight into how MV accomplishes trans-neuronal spread, a process that we know to be unique and distinct from the “classical” spread of MV through non-neuronal cells.Top
Mechanisms of Cytokine-Mediated Viral Clearance from Neurons
In order to examine interactions between virally-infected CNS neurons and the immune system, our laboratory uses a transgenic mouse model of neuron-restricted measles virus (MV) infection. These transgenic mice (NSE-CD46 mice) express CD46, a MV receptor, under the control of the neuron-specific enolase (NSE) promoter. Work from our laboratory demonstrated that adult NSE-CD46 mice successfully clear MV from the brain through a T-cell dependent mechanism involving interferon-gamma (IFNγ). Moreover, CD46+ hippocampal neurons explanted from NSE-CD46 embryos control MV infection when pretreated with IFNγ. However, when the activation of the canonical JAK-STAT1 pathway was examined in neurons, STAT1 expression and activation (phosphorylation) was severely diminished and delayed in comparison to fibroblast controls. This suggests that neurons may utilize alternative signaling mechanisms to cope with viral infection and IFNγ treatment. Thus, my first aim sought to determine the role of STAT1 in IFNγ signaling and viral control in primary neurons, and to define alternative pathways that are triggered by IFNγ in neurons.
To determine if STAT1 is required for control of MV infection in neurons, CD46+ and CD46+/STAT1-KO neurons were infected
with MV (MOI 1.0) in the presence of IFNγ (100 U/ml; 24h before infection.) CD46+ neurons effectively control MV infection when treated with IFNγ, as measured by total MV protein levels by western blot. CD46+/STAT1-KO neurons, which lack functional STAT1, control MV to similar levels as CD46+ neurons (Figure: STAT1 is not required for controlling MV infection in primary neurons). This suggests that STAT1 is not required for successful control of MV infection in primary neurons upon IFNγ treatment, and that other signaling pathways may be significant in IFNγ-mediated neuronal signaling.
To determine if alternative signaling pathways are activated in neurons upon IFNγ treatment, CD46+ hippocampal neurons were treated with IFNγ over a 24-hour time course and analyzed for activation of AKT and Extracellular Regulated Kinase-1/2 (ERK-1/2), which have been shown to activate ISG expression in other models. IFNγ caused rapid and robust phosphorylation of AKT and ERK-1/2 in hippocampal neurons, but not in control embryonic fibroblasts. This analysis suggests that neurons can readily respond to IFNγ, but may activate different downstream signaling cascades from cycling cells.
IFNγ treatment is associated with modification of glutamate receptors, dendritic beading and apoptosis in some primary neuron systems, though we have demonstrated activation of cell survival kinases in our model system. Given the extended exposure to IFNγ over the course of the infection experiments and its potential neurotoxicity, we have analyzed the levels of dendritic beading by MAP-2 staining in CD46+ hippocampal neurons over a 72-hour time course to assess neuronal damage. IFNγ-treated cultures exhibited similar levels of dendritic beading as control cultures (data not shown), suggesting that the prolonged IFNγ exposure is not neurotoxic to hippocampal neurons.
To determine the role of STAT1 in MV clearance from CNS neurons, NSE-CD46+ and CD46+/STAT1-KO adult mice were infected intracranially with MV (2x10^4 PFU) and monitored for signs of clinical illness and mortality. CD46+/IFNγ-KO mice, which lack IFNγ, and CD46+/RAG2-KO mice, which lack B and T-cells, were included as controls. All of the NSE-CD46+ mice survived MV infection (Figure: STAT1 is not required for surviving MV infection in the CNS). Similar to previous studies, half of CD46+/IFNγ-KO mice succumbed to MV infection, indicating the importance of IFNγ in controlling MV in the CNS. The CD46+/RAG2-KO mice were the most susceptible to infection, with 90% of the adult mice dying by 30 dpi. In contrast, the majority of CD46+/STAT1-KO mice survived the infection (70% survival.) The CD46+/STAT1-KO mice that succumbed to infection died early in infection (2-4 dpi), suggesting a possible role for IFNα/β signaling in controlling the initial infection. Moreover, there was no evidence of MV infection in the brains of NSE-CD46 and CD46+/STAT1-KO by 30 dpi, as analyzed by quantitative RT-PCR for MV transcripts or by immunohistochemistry for MV antigens. These results suggest that STAT1 is not required for successful MV clearance from CNS neurons and that other IFN pathways may be utilized by CNS neurons.Top
Pathogenic Consequences of Aberrant Immune Response Targeting During Concurrent Viral Infections
It has been hypothesized that viruses may play a role in central nervous system (CNS) diseases of unknown etiology, yet to date identification of such viral triggers has been largely unsuccessful. To elucidate whether a peripheral infection can contribute to neuropathogenic disease, we have developed a mouse model in which immune cells generated to a peripheral infection are recruited to the CNS, where they subsequently mediate immunopathogenesis. Specifically, NSE-CD46+ transgenic mice are infected with CNS-restricted measles virus (MV) and lymphocytic choriomeningitis virus (LCMV), which is restricted to the periphery. While infection with either virus alone has no effect, concomitant infection results in disease in approximately half of the doubly infected mice. The posture of co-infected mice upon death is indicative of immunopathology, resembling that seen in mice following intracerebral challenge with LCMV, where mortality is CD8+ T cell dependent. Characteristic signs of lymphocytic choriomeningitis, such as unilateral pupillary dilation and compression of the brainstem and cerebellum against the skull, were noted in symptomatic, doubly infected mice. In addition, loss of blood-brain barrier (BBB) integrity and increased edema, likely caused by leakage of cerebrospinal fluid (CSF) from the ventricles into the parenchyma, were observed. Interestingly, disease in the double infection model correlates with a >10-fold increase in the number CD8+ T cells in the CNS. Furthermore, GP33 and NP396 tetramer staining suggests that a substantial proportion (~40%) of these infiltrating CD8+ T cells are LCMV specific, despite the absence of this virus from the CNS.
These results suggest a potential role for mis-recruitment of the peripheral immune response in triggering novel CNS pathologies. Ongoing studies will elucidate the mechanistic basis of this immunopathogenesis and examine whether aberrant trafficking of activated bystander memory cells can facilitate a similar effect to that observed with contemporaneous infections.Top