BIOLOGICAL ASPECTS OF INTRACELLULAR STAGES OF VIRUS GROWTH 127 



soluble antigen have usually been made with sera from adults convalescent 

 from influenza, although such sera are unsuitable for tests with viral antigen 

 on account of their broad reactivity with different strains of influenza A virus. 

 In order to measure specific viral antigens most workers use sera of con- 

 valescent or immunized animals which have not had previous experience 

 with influenza virus and therefore respond to infection or immunization in a 

 more specific way. When sera from infants infected (presumably for the first 

 time) with influenza virus were tested with soluble antigens prepared from 

 different influenza A strains, the soluble antigens showed some strain 

 specificity (Grist, 1957). There is therefore a real difficulty in defining the 

 influenza soluble antigen precisely. It is normally taken to mean an antigen 

 of sedimentation constant 30 S, present in infected tissues, and showing the 

 broad serological reactivity of the virus serotype. It must be emphasized 

 however, that the term soluble antigen may have been used by different 

 workers to describe different things; this is especially true when studies with 

 human convalescent serum are compared with investigations using animal 

 sera reacting with soluble antigen. 



Ho vie (1948) found that in cells infected with influenza virus, soluble 

 .antigen could be detected at the end of the eclipse period and before the 

 formation of mature virus. Subsequently, Hoyle (1950) showed that, on 

 treating virus elementary bodies with ether, soluble antigen was liberated, 

 .and he therefore suggested that the soluble antigen represented the essential 

 form in which the virus multiplied within the cell. This is an attractive theory 

 which attempts to draw an analogy between the influenza soluble antigen and 

 the nucleic acid of bacteriophages and tobacco mosaic virus. At the moment, 

 the evidence concerning this theory is as follows: 



1. Soluble antigen is detectable in tissues within 3 hours of inoculating 

 influenza virus, whereas fully infective virus is not formed until the fourth 

 to the fifth hour (Henle and Henle, 1949). However, while this finding has 

 been generally confirmed, there is one reservation which should be made. In 

 experiments on the time of appearance of soluble antigen, large virus inocula 

 are required to produce measurable yields of soluble antigen, and under these 

 conditions multiple infection of cells occurs and incomplete virus is formed 

 (see Section VI). Hence, to overcome this objection, the rise of viral infecti- 

 vity must be studied after the use of small inocula; this means that the 

 development of soluble antigen and infective virus have not been observed 

 under comparable conditions. There are a number of known instances with 

 other viruses in which newly formed virus has been shown to appear earlier 

 with large inocula, e.g., Dulbecco and Vogt (1954), so it is possible that the 

 earlier appearance of the influenza soluble antigen may be more apparent 

 than real. Recently, Ledinko and associates (1957) have described a rise in 

 infectivity 21 to 2| hours after infection of chick embryo lung cells with 



