SEROLOGICAL VARIATION 267 



expression of the total antigenic complement. The differences seem to be more 

 quantitative than qualitative alterations in composition. 



2. Induced Variation 



a. In Eggs. Several studies have reported the alteration in antigenic 

 behavior of strains grown in an environment of antibody (Archetti and 

 Horsfall, 1950; Isaacs et al, 1952; Burnet and Lind, 1954; Edney, 1957). 

 Horsfall (1952) has especially commented that in their study the results 

 were obtained when large amounts of virus were introduced, just as is true of 

 production of "incomplete" virus. On the other hand, Magill and Jotz (1952) 

 have recalled the unsuccessful efforts made earlier to induce such changes with 

 immune serum in tissue culture. A number of the variations have apparently 

 stabilized so as to persist in the absence of immune serum. Nevertheless, they 

 are of limited character and, as suggested above, may be quantitative rather 

 than truly constitutional. 



6. In Mice. Gerber and associates (1955, 1956) detected four successive 

 generations of antigenic variants of PR8 strain, each derived from the 

 previous one by serial passage in the lungs of mice immunized with the 

 homologous agent. The variants exhibited progressively decreasing reactivity 

 with the parent PR8 antiserum, while retaining the ability to elicit antibody 

 to the original PR8 strain and to their respective predecessors. These were 

 stable mutants. Antibody-absorption tests revealed that they differed from 

 the original in the appearance of previously unrecognized antigenic compo- 

 nents. Nevertheless, vaccination of mice with PR8 strains protected against 

 the variants. Magill (1955) also reported the recognition of variants developing 

 in immune mice; the nature of these he considered to be antigenic rearrange- 

 ment. An important feature of these observations again is to emphasize that 

 the shifts occur within limited boundaries of related structures in association 

 with immunological pressures. 



3. Phase Variation 



One of the difficulties in evaluation of serological variations which can be 

 demonstrated is the presence of other substances in serum which can interfere 

 with serological activity of the virus and simulate antibody. One of these 

 associated with mucopolysaccharides of serum has been extensively studied 

 (Gottschalk, 1954, 1957). It was shown (Francis, 1947) that if influenza virus 

 is heated, its capacity to combine with erythrocytes is inhibited to high titer 

 by normal serum. Some strains, especially when first isolated in eggs, are 

 especially susceptible to normal serum inhibitors so that their antigenic 

 pattern is difficult to determine promptly. This exaggerated tendency to 

 combine with inhibitor may be a reflection of epidemic capacity, indicating 

 a high ability to combine with receptors of any kind. This effect may be 



