PATHOGENESIS OF NEUROTROPIC VIRAL INFECTIONS

GLENN F. RALL, Ph.D., Associate Member; Adjunct Professor of Microbiology, University of Pennsylvania

Our laboratory's primary objective is to uncover basic paradigms that govern viral infection and the resulting immune response within the central nervous system (CNS). The basis for our interest is two-fold. First, a large number of viruses of clinical significance to humans can infect cells within the brain and spinal cord, including the herpes viruses, HIV-1, measles, rabies and polio. All of these viruses can establish long-term or chronic infections in the CNS, some of which result in neural damage after a protracted period of infection. The basis of these CNS diseases is poorly understood. Second, from an immunological perspective, the CNS is distinct from peripheral tissues. The restriction of immune function within the brain, due to the presence of the blood-brain barrier, a lack of major histocompatibility complex (MHC) expression on resident CNS cells, and an absence of lymphatic drainage, has led to the presumption that the brain is a site of immune privilege. However, it is also clear that activated leukocytes and immune mediators can infiltrate the brain, and that inflammatory responses are associated with numerous CNS disorders, including many neurotropic viral infections. While the antiviral immune response may help control infection in the CNS, often it is this response, rather than the infection alone, that results in CNS disease. Thus, how the immune response functions in the virally infected brain remains unclear.

To understand how the balance between neuroprotection and immunopathogenesis is regulated in the CNS, we are addressing three main experimental questions: 1) How do viruses gain access to the CNS from the periphery? 2) How are neurotropic viruses transmitted within the brain? 3) What role, if any, does the host immune response play in the control of such infections? To address these issues, we use a combination of transgenic mouse and primary cell culture models to study the infections caused by measles virus (MV) and lymphocytic choriomeningitis virus (LCMV). By understanding the factors that contribute to CNS diseases caused by neurotropic viral infections, we will be in a more informed position to develop strategies to prevent or resolve such infections.

MV is among the most contagious infections of humans, causing greater than one million infant and child deaths annually despite the existence of an effective vaccine. Subacute sclerosing panencephalitis (SSPE) is a rare CNS complication of MV infection in which an extensive loss of cerebral cortical functions ultimately leads to coma and death. The basis for this MV-associated CNS disease is unclear, and the role that the host immune response plays in either neuroprotection or disease is not understood. Thus, a transgenic mouse model was developed in which neurons of the mouse CNS express the human MV receptor, CD46. Our previous work demonstrated that CNS neurons of CD46⁺ transgenic mice can be infected with MV, and that CNS disease (including blood-brain barrier permeability alterations and hydrocephaly, both hallmark features of human SSPE) and eventual death occur in neonatal mice, but not in adult mice (Rall et.al., Proc. Natl. Acad. Sci. 94:4659, 1997).

While sickness and death following MV infection are age-dependent, T lymphocytes infiltrate the brain in response to inoculation at all ages. To determine the role of lymphocytes in disease progression, CD46⁺ mice were backcrossed to T and B cell-deficient RAG-2 knockout (KO) mice. The resulting adult or neonatal CD46⁺/RAG-2 KO mice were infected with MV. The lymphocyte deficiency did not affect the outcome of disease in CD46⁺/RAG-2 KO neonates; however, 71% of adult CD46⁺/RAG-2 KO mice were susceptible to both neuronal infection and CNS disease compared to 0% of immunocompetent littermates (Table 1). These results indicate that CD46-dependent MV infection of neurons, rather than the antiviral immune response in the brain, produces neurological disease in this model system, and that immunocompetent adult mice, but not immunologically compromised or immature mice, are protected from infection.

To determine the contribution of T cell subsets in neuroprotection, CD46+ mice were also backcrossed to mice deficient in T cell subsets expressing CD4 (CD4 -/-) or CD8 (b2-microglobulin [b2mic -/-]). While immunocompetent adult mice were resistant to infection, 58% of adult CD46+/CD4-KO mice and 39% of adult CD46+/b2mic-KO mice developed signs of CNS disease, indicating that both subsets contributed to protection (Table 1). The absence of NK cell cytotoxicity did not greatly influence the response to neuronal measles virus infection. Increased sickness correlated with extensive infection of both the brain and spinal cord, and the absence of one T cell subset did not reduce the CNS infiltration of the other subset. C-C chemokine mRNA expression was induced by MV infection in vivo as well as in primary transgenic mouse neuronal cultures, suggesting that MV infection of neurons may trigger the production of chemotactic factors to independently recruit CD4+ and CD8+ T cells into the infected brain.

MV recovered from brains of SSPE patients contains sequence mutations and is defective in extracellular virus production, suggesting that the mechanism of viral spread in neurons is distinct from that in other cell types. To study MV transmission between neurons, MV-Edmonston infection of cultured NT2 neuroblastoma cells and primary transgenic mouse neurons, both of which express the MV receptor, CD46, was compared to infection of permissive fibroblasts. Like Vero and HeLa cells, the percentage of MV-immunopositive neurons increased over time in a focal pattern, suggesting that replication and interneuronal spread had occurred. In contrast, infected neurons did not form syncytia, and extracellular virus production was reduced at least 1000-fold. These differences were not due to a lack of cell division in neurons; infected fibroblasts cultured with mitotic inhibitors were not impaired in either syncytia formation or extracellular virus production, whereas both primary neurons and differentiated NT2 neurons produced little infectious virus (Figure 1). Electron microscopy studies showed nucleocapsid alignment and envelope protein expression at the neuronal cell surface, including synapses, but few mature buds were visible. These results suggest that the neuronal environment prevents normal mechanisms of MV spread between neurons while allowing an alternate mechanism, possibly via synaptic transmission.

A COMPROMISED IMMUNE RESPONSE FAVORS A PERSISTENT NEURONAL INFECTION. VAUGHN, LAWRENCE, RALL, in collaboration with DAVÉ,^§ KAPPES^§

For a virus to establish a persistent infection, it must be noncytolytic and must not activate an effective host immune response. One strategy for persisting RNA viruses to evade immune recognition and elimination is to reside in cells that cannot be seen by the host response, such as the class I major histocompatibility complex (MHC)-negative neurons of the CNS. In order for this strategy to be successful, however, the virus must gain access to neurons before the host immune response can clear the infection; consequently, the kinetics of neuroinvasion and the efficiency of immune induction must play critical roles in the pathogenesis of neurotropic viral infections.

To test this hypothesis, we examined the time course of LCMV infection following intracerebral inoculation of wild type mice and mice that exhibit impaired T cell responsiveness due to the lack of the CD3 d subunit of the T cell signaling complex (CD3 d-/- mice). These mice represent a unique model for organisms with partially compromised or immature T cell compartments. In all mice, infection was localized to cells of the meninges, ependyma and choroid plexus during the first week post-infection. Immunocompetent mice succumbed to LCMV-induced, CTL-mediated neuropathology. In contrast, CD3 d-/- mice survived with infection shifting from the meninges and ependyma to neurons within the brain parenchyma between 10 and 30 days post-inoculation. At the same time, CD3 d-/- mice developed a delayed T cell response, which partially suppressed virus replication in peripheral tissues, but not in the CNS; this was likely because the virus was sequestered in a cell type that is protected from CTL recognition. Our results suggest that a T cell-compromised or developing immune response may allow neurotropic viruses to access the immune-privileged CNS and establish persistent infections.

Lawrence, D.M.P., Vaughn, M.M., Belman, A.R., Cole J.S., Rall, G.F. Immune response -mediated protection of adult but not neonatal mice from neuron-restricted measles virus infection and central nervous system disease. J. Virol. 73:1795-1801, 1999.

Pugh, J.C., Guo, J.T., Aldrich, C., Rall, G., Kajino, K., Tennant, B., England, J.M., Mason, W.S. Aberrant expression of a cytokeratin in a subset of hepatocytes during chronic WHV infection. Virology 249:68-79, 1998.

RALL, G.F. CNS neurons: the basis and benefits of low MHC expression. Curr. Top. Microbiol. Immunol. 232:115-134, 1998.

^§   Fox Chase researcher

^a   A.L. Lehman: Present address--University of Pennsylvania, Philadelphia, PA 19104

Illustrations or unpublished data in these reports should not be used without permission of the author.

Fox Chase Cancer Center

Scientific Report 1998