Clinical sequelae of human infection by neurotropic viruses have been recognized for nearly 4,000 years. A funeral stela of a XIXth dynasty Egyptian priest showed the effects of prior poliomyelitis.’The epitome of classic clinical and pathologic studies of neurotropic viruses can be found in Pasteur’s work on rabies2 Until recently, it was not possible to explain the fundamental clinical properties of viruses at the level of the viral gene. Nowhere is this difficulty better exemplified than in the problems of neurotropism and the spread of virus within the nervous system.

It has been known for over a century that the striking clinical variability in infection by polio or rabies implies selective infection of discrete neuronal populations. Using experimental reovirus infection, it has been possible to establish the molecular and genetic basis of viral neurotropi~ m.”.~ Reoviruses are nonencapsulated viruses 600 to 900. & in diameter, with icosahedral symmetry. 5 “Reo” is an acronym for respiratory and enteric orphan virus and reflects the original sites of isolation of these viruses as well as the initial lack of associated disease (hence “orphan”). The viral core contains 10 unique segments of double-stranded RNA (dsRNA) that can be subdivided into three classes based on molecular weight. Each dsRNA segment encodes a single strand of messenger RNA that is translated on host ribosomes into a unique single polypeptide. Two of the small dsRNA segments (S1 and S4) and one medium segment (M2) code for structural polypeptides (o1, a {, and pic), which together form the outer capsid The three distinct serotypes (types 1, 2, and 3) of reovirus identified by hemagglutination inhibition and antibody neutralization tests also differ in electrophoretic mobility of the RNA segments and proteins. Stable recombinants containing specific desired gene combinations can be isolated and perpetuated unchanged