GENETICS OF SOMATIC CELLS 447 



level. 775 These developments have also greatly influenced thinking about viral 

 tumors, and many attractive models have been offered in which the possible mechanism 

 of viral oncogenesis is considered in the light of the information obtained about such 

 phenomena as transduction and lysogenic conversion. 810, 811> 1380 It has been re- 

 peatedly emphasized that the contrast between the somatic mutation and the viral 

 causation theory of cancer is probably more semantic than substantial. As pointed 

 out by Luria, 811 the concept of infective heredity is mainly based on the following three 

 groups of findings: " the central and often exclusive role of viral nucleic acid in initiating 

 virus infection; the interactions between viruses and genetic constituents of the cell; 

 and the viral control of cellular functions through determination of the structure of 

 specific proteins." In Luria's view, viral infection may be considered as a class of 

 cellular mutations, in which the primary change, the entry of the viral genome, is a 

 genetic change. 



Perhaps most suggestive of an analogy with viral tumors are the cases of lysogenic 

 conversion in which the genome of the converting phage contains determinants that 

 control cellular properties, such as surface antigens, which have no apparent relation 

 to the viral function of the phage. Cases at least superficially similar to this situation, 

 in which viral onocogenesis has led to a new phenotypic cellular property, in addition 

 to the characteristics recognized as related to the neoplastic change itself, have been 

 discussed recently by Luria. 811 He pointed out that such cases deserve particular 

 attention in view of the fact that the amount of genetic information in many animal 

 viruses must be quite limited and therefore any new cellular function that can be shown 

 to be directly controlled by viral genes has "a significant chance of being the key 

 function in the tumoral transformation." One such case is the appearance of arginase 

 in cells infected by Shope papilloma virus. 1069 This enzyme is absent from normal 

 rabbit epidermis but present in papilloma cells and in the cancers that originate from 

 them. The host animals contain a precipitin in their serum, reacting with the tumor 

 arginase that is absent from the control animals. Another analogous case has been 

 described recently by Zilber 1469 who found a specific antigen in cells of the Rous 

 sarcoma, not present in the Rous virus itself or in methylcholanthrene-induced sarco- 

 mas in chickens. A similar situation is suggested by the recent findings of Sjogren 

 et al., 1215 who showed that mice preimmunized against the polyoma virus were resistant 

 against the transplantation of polyoma-induced tumors that have arisen in isologous 

 mice and were capable of growing in untreated isologous hosts. 



This is not the suitable place to enter into lengthy hypothetical discussions about 

 the relationship between the virus and the cellular genome in virus-induced tumors. 

 Only one particular view will be quoted, which has been put forward recently by Vogt 

 and Dulbecco, 1350 who studied the interactions of polyoma virus with cells of mouse 

 and hamster in vitro. They found that the virus could give rise to two types of virus- 

 cell interaction : a cytocidal interaction, leading to extensive viral synthesis and cellular 

 degeneration, and a moderate interaction leading to the transformation of the cells into 

 neoplastic cells, usually unable to produce detectable virus and resistant to superinfection 



