IMMUNOLOGICAL METHODS IN THE STUDY OF VIRUSES 541 



Dulbecco (1952) provided an opportunity of applying the same principles 

 that were used in pock counts on the chorioallantois to a more elegant and 

 controllable system. The paper by Dulbecco and associates (1956) is the most 

 important contribution on the theoretical side that has yet appeared. They 

 were concerned with both WEE and poliovirus and, although the results 

 obtained were generally similar in quahty to those reported by Burnet et al. 

 (1937) on the chorioallantois, their mterpretation was radically different. 

 All who had previously discussed the matter had agreed that a reversible 

 reaction between virus and antibody must be involved. Dulbecco et al. 

 (1956) concluded: (1) Neutralization is a direct consequence of the com- 

 bination of the virus particle with antibody molecules; (2) the kinetics of 

 neutrahzation in presence of antibody excess are of the first order; and (3) 

 the rate of neutralization is linearly dependent on the concentration of anti- 

 body. From (2) and (3) it is deduced that attachment of a single molecule of 

 antibody is sufficient to inactivate a particle; (4) the virus-antibody com- 

 plexes formed are very stable; (5) virus particles of the types used could 

 absorb up to 15 equivalents of antibody; and (6) the characteristics of the 

 process are independent of the cell system used for assay. 



Dulbecco's view may be summarized simply by saying that when animal 

 viruses are handled like bacteriophages the virus-antibody reaction has the 

 same characteristics. Most readers of their paper will, however, feel that too 

 little attention has been given to providing an acceptable explanation of the 

 residual infectivity which, like all other experimenters, Dulbecco et al. (1956) 

 find in the presence of a high antibody concentration. With WEE the 

 "persistent fraction" diminishes regularly with increasing serum concentra- 

 tion, while with poliovirus it remains approximately constant over a fairly 

 narrow range of serum dilutions. To account for this, it is assumed (7) that 

 each virus preparation contains a fixed fraction of unneutralizable virus 

 particles. (The quality is not genetic and is assumed to be due to non- 

 hereditary differences in antigenic constitution.) The further fact that addi- 

 tion of inactive virus to a neutralized mixture releases infective virus is 

 rendered compatible with the claim that ^'irus-antibody union is essentially 

 irreversible by assuming (8) that transfer of antibody is possible only by 

 thermal collision between inactive virus, and antibody-coated active virus 

 particles. 



Fazekas de St. Groth and associates (1957) have recently provided a 

 serious criticism of Dulbecco's theoretical treatment, holding that the data 

 presented in the paper by Dulbecco et al. (1956) is better interpreted on the 

 assumption of fully reversible reactions. (I am indebted to Dr. Fazekas de 

 St. Groth (1958) for the opportmiity to see a prepublication draft of this 

 paper, from which the following summary is derived.) 



(a) It is assumed that dissociable complexes are formed by union between 



