VARIATION IN VIRULENCE 243 



occurring in the intestinal wall, the local lymph nodes, some central extra- 

 neural focus, and the central nervous system. Within the central nervous 

 system there may be further barriers to the spread of virus, for histological 

 examination of orally infected monkeys and chimpanzees has shown that 

 asymptomatic infection of the nervous system may occur in some animals, 

 whereas in others there may be extensive spread through the spinal cord and 

 the brain. 



Sabin's extensive studies on variation in the neuropathogenicity of the 

 polioviruses show clearly the great complexity of this property. The impossi- 

 bility of subjecting it to adequate genetic analysis at the present time is 

 further exemplified by the observation that at least three different genetic 

 factors are involved in the determination of neuropathogenicity (Vogt et al., 

 1957). 



Variants of poliovirus have been described which illustrate the failure of 

 virus to multiply in one or other tissue of the sequence illustrated in Fig. 1. 

 First, a high degree of neuropathogenicity of intracerebral inoculation may 

 be dissociated from the ability to establish infection by feeding (Melnick, 

 1951; Sabin, 1956). Second, variants which are able to multiply in the 

 alimentary mucosa (and provoke antibody formation) may fail to be 

 immunogenic after intramuscular inoculation, i.e., such variants may fail to 

 multiply in lymphoid tissue (Sabin, 1955a). Third, variants incapable of 

 producing either paralysis or lesions after inoculation of large doses into the 

 lumbar cord of chimpanzees will nevertheless multiply extensively in their 

 alimentary tract (Sabin, 1955b,c). 



During the last few years it has become apparent that the polioviruses are 

 members of a large group of viruses which normally parasitize the enteric 

 tract of man — a group now designated as the "enteroviruses." From the point 

 of view of survival in nature, all that is required of such viruses is that they 

 should be able to infect the cells of the alimentary tract and be excreted in 

 the feces; this appears to be the limit of the activity of most members of the 

 group. The aim of those who seek attenuated variants of the polioviruses for 

 use as oral vaccines is essentially to obtain virus strains with the limited 

 invasive power of most enteroviruses but the same antigenic constitution as 

 those which occasionally invade the central nervous system of man. The 

 major difficulty appears to be the occasional acquisition by attenuated 

 variants of some degree of neuropathogenicity after multiplication in the 

 intestinal tract of man (Sabin, 1955c; Dick and Dane, 1957). The suggestion 

 made by Dick and Dane that, to counteract this risk, the virus should not be 

 transmissible from vaccinated to nonvaccinated people may well be incom- 

 patible with the requirement that the oral vaccine should be immunogenic. 



It is perhaps relevant that involvement of the central nervous system is of 

 no evolutionary significance for most viruses, whereas in the exanthemata 



