INTERFERENCE BETWEEN ANIMAL VIRUSES 159 



This restriction introduces into the concept of interference a complex and 

 obscure parameter. Attempts to analyze the essence of virulence lead to the 

 conclusion that it is an expression of multiple genetic factors. Recombination 

 experiments with influenza viruses (Burnet, 1955, p. 431; Lind and Burnet, 

 1957b; Gotlieb and Hirst, 1954) indicate that the dosage of such factors is 

 highly variable. Hence the pheDotypic expression of the entire complex, 

 virulence, is easily lost in recombinant progeny. The lability of this character 

 has obvious relevance to our problem. Many studies of interference concern 

 protection of the host or of cells against a virulent virus resulting from 

 simultaneous infection with an avirulent variant of the same or a related 

 agent. In interpreting such situations, we must consider the possibility that 

 protection might in reality be due to. loss of virulence in combinant viral 

 progeny rather than to interference in the sense of mutual exclusion. The 

 realization of this conflict will be reflected in the following pages, specifically 

 in Section III, A, 2, b. 



2. Mutual Exclusion and Nonspecific Inhibition of Viral "Toxicity" 



The term "viral interference" conventionally has the connotation of mutual 

 exclusion (Delbrtick, 1945; Dulbecco, 1952) which, in turn, implies that 

 superinfecting viral particles are somehow prevented from entering and 

 multiplying in an already infected cell. In the case of bacterial viruses, 

 however, a cell resistant to superinfection by virtue of mutual exclusion is 

 not protected against "lysis from without" by an appropriate phage at high 

 multiplicity (Doermann, 1948). In recent years, much evidence has accumula- 

 ted for "toxic" cell damage by animal viruses (Cox, 1953) which is often 

 independent of multiplication. What, if any, relation does protection or lack 

 of protection against such effects have to interference? 



The problem here is that protection can be achieved not only by pretreat- 

 ing the host with virus (Prince and Ginsberg, 1953; Fong et ah, 1953; Wagner, 

 1953; Groupe and Dougherty, 1956; Dougherty and Groupe, 1957; Manire, 

 1957; Herrmann et al., 1955; Khoobyarian and Walker, 1957), but also with 

 nonviral agents, such as Xerosin, bacterial vaccines, or heated receptor- 

 destroying enzyme (RDE) (Groupe et al., 1954; Groupe and Herrmann, 1955; 

 Dougherty and Groupe, 1957; Wagner, 1953; Fong et al., 1953; Ginsberg, 1955; 

 Khoobyarian and Walker, 1957). Protection against toxic viral manifestations 

 can be achieved without affecting viral multiplication (Groupe et al., 1954b) 

 or "incomplete" virus production (Frankel and Schlesinger, 1952; Ginsberg, 

 1955). Ginsberg has studied the effect of Xerosin on pneumonitis induced in 

 mice by the CAM strain of influenza virus. The agent does not inhibit specific 

 virus-induced injury to susceptible cells, but does prevent secondary reactions 

 leading to pulmonary consolidation and death. 



