THE MORPHOLOGICAL APPROACH 85 



B. Filament Formation 



Since more is known about the surface activity, however, we will discuss 

 it first. The presence of a variable though usually small number of typical 

 spherical forms of the virus along the filament and usually at the tip has 

 been described by several authors (Murphy et al., 1950; Bang and Isaacs, 

 1957; Wyckoff, 1953; Archetti, 1955) and early suggested an association 

 between the two (Figs. 5, 6, 7). On the basis of an earlier observation, Robinow 

 (1950) proposed that these filaments arose from the edge of the cell. There 

 followed a series of observations on tissue cultures (see Table I) which 

 showed that a profusion of filaments may occur when previously infected 

 cells are grown in tissue culture, and that virus action was directly related to 

 the surface of the cell. This relationship was established by tissue culture 

 studies of both influenza (Murphy et al., 1950; Murphy and Bang, 1952) and 

 Newcastle disease virus (Bang, 1953a) (Figs. 8, 9, 10). Material taken at inter- 

 vals following infection with three members of the myxovirus group — 

 influenza (Bang and Isaacs, 1957; Morgan et al., 1956a) Newcastle (Bang, 

 1953b) and fowl plague, and sectioned (Hotz and Schaefer, 1955) showed a 

 profusion of filaments put forth at the surface of the cell during the later 

 stages of infection. Major surface changes may occur without apparent change 

 in the structure of the underlying cytoplasm. These cells normally have a 

 number of microvilli on their surface (Borysko and Bang, 1953; Bang, 1955a; 

 Morgan et al., 1956a). This, combined with the fact that cells frequently form 

 arrays of variable extrusions when damaged, suggested that this process was 

 nonspecific (Hoyle, 1950). However, the presence of spherical morphological 

 units within the filament, of red cell-agglutinating activity along the filament, 

 and the great length and relatively uniform diameter of the filaments indicate 

 a fair degree of specific viral activity in the filament (Bang and Isaacs, 1957). 

 Recent enzymatic analysis of these filaments by Valentine and Isaacs (1957a) 

 has shown that the filaments developed rows of spheres along the long axis 

 when treated with acid and that these spheres were completely digested with 

 trypsin. On the other hand, the separate spherical forms contained trypsin- 

 resistant rings, tentatively identified as ribo-nucleoprotein (Valentine and 

 Isaacs, 1957b). 



Filament formation is particularly characteristic of cells infected with 

 recently adapted strains of influenza, and very similar long cellular extrusions 

 are found with infection with the mild or vaccine strain of Newcastle disease 

 virus. 



In contrast, the virulent strain of virus destroys the cell, and it is assumed 

 that a virus is liberated as the cell boundaries disintegrate around it. The two 

 distinctive devices by which avirulent and virulent viruses are released from 

 a parasitized cell would inherently affect the pathogenesis of the disease in 

 the host. However, it should be emphasized that no particles identifiable as 



