INTERFERENCE BETWEEN ANIMAL VIRUSES 185 



interference was effective 6 days after inoculation of inactivated virus 

 (Henle, 1950). Since basically nothing is known about the nature of the 

 association of these inactivated agents with their host cells, one cannot say 

 what factor is responsible for the ultimate cessation of interference. On the 

 other hand, the persistence of exclusion of superinfecting viruses from long- 

 term "carrier" cells (see Section III, C, 2) suggests that interference is 

 effective as long as the interfering agent remains actively associated with the 

 cell. This statement implies nothing in regard to any possible analogy of such 

 systems to lysogeny which has not been proved in a single case. 



3. Efficiency — "Overcoming" of Interference 



Baluda (1957) finds that "there exists for each interfered cell a small 

 probability that it can be superinfected by any given active virus particle. 

 From the results obtained, either the maximum value of this probability is 

 very small for all the cells, i.e., less than 4 %, or it has a value of 1 for a small 

 fraction of the cells only." This statement implies, of course, that the chances 

 of "overcoming" interference increase with the multiplicity of superinfecting 

 viral particles. 



The corresponding situation for influenza virus has not been as clearly 

 denned. As indicated, establishment of solid interference requires up to 16-24 

 hours. Whether the gradually decreasing yield of progeny of superinfecting 

 virus up to that time is due to "bypassing" (Fazekas de St. Groth et al., 1952) 

 or to reduced yield from all cells is uncertain. Evidence in favor of the former 

 view may be invalidated by use of maximally effective interfering virus 

 (Henle and Paucker, 1958). In contrast to the NDV system, increased multi- 

 plicity of superinfecting virus does not significantly increase the final yield 

 (Henle and Henle, 1944b; Fazekas de St. Groth et al., 1952). A complicating 

 factor, already mentioned, is the theoretical possibility that some apparent 

 "overcoming" in homologous influenza systems may in fact be due to reactiva- 

 tion and recombination between interfering and superinfecting virus. 

 Isaacs and Edney (1951a) observed strain-specific differences in the ease with 

 which different variants of type A strains could "overcome" interference 

 induced by a standard dose of heated LEE. They thought at the time that 

 this indicated that complete interference could not necessarily be accounted 

 for by the assumption that 1 particle of heated LEE per cell assured inter- 

 ference. This is perhaps one situation in which further studies on interferon 

 (Isaacs et al., 1957) will clear up apparent quantitative discrepancies. 



V. Mechanisms of Interference 



A uniform approach to the theory of interference is impossible. Instinct 

 tells us that the mechanism whereby influenza virus interferes with homologous 



