454 GENETICS OF SOMATIC CELLS 



and the C3H mice in which they are actually tested. These differences may be expressed 

 to different degrees in the two lines, representing another case of isoantigenic variation, 

 well demonstrated to occur in systems of transplanted tumor in vivo. Alternatively, 

 the two lines may have developed an antigenicity of their own, not yet present in the 

 C3H mouse from which their common ancestor had been derived, in analogy with 

 individual antigenicity of methylcholanthrene-induced sarcomas previously discussed. 



Sanford et a/. 1151 ' 1152 carried out critical experiments to decide the question of 

 antigenic difference between the lines versus a difference in malignancy. They found 

 that in irradiated hosts with homograft reaction paralyzed, the two lines differed 

 significantly with regard to the latent period elapsing before the appearance of palpable 

 tumors. Cells of the high line proliferated more rapidly when first implanted into 

 mice than low-line cells. In contrast, both lines showed the same growth rate when 

 compared in vitro. There was also a difference between the ability of the two lines to 

 induce vascularization: high-line cells became more rapidly vascularized than equal 

 numbers of low-line cells. It was concluded that the difference between the two 

 lines was not due to the fact that one was antigenic and the other was not, since both 

 were antigenic, but to the fact that the high-line cells could produce an established tumor 

 before resistance developed in the host. This in turn depended on a more rapid rate of 

 proliferation of the high-line cells in vivo, presumably because of a difference between 

 the two lines in their response to some growth-controlling factor in the host. 



Critical studies of this and similar type are urgently needed to clarify this field. 

 Whenever possible, it will be highly desirable to exclude or minimize the artifact of 

 isoantigenic divergence between cells and host strain. This can be best insured by 

 the use of cells derived from critically tested homozygous strains of animals. Exchange 

 of skin grafts is probably the best available test for isogenicity. In addition, the time 

 elapsing between the first explantation of the tissue and the retransplantation experi- 

 ment should be restricted to the shortest possible period, preferably not more than one 

 or two years. If this is not feasible, the genotype of the original donor animal may be 

 partially reproduced by combining modern techniques of frozen storage of spermatozoa 

 with artificial insemination. Whatever the genotype of the recipient female, if the 

 donor male was identical or isogenic with the animal from which the strain of cells 

 had been derived and truly homozygous, the offspring will contain all histocompati- 

 bility factors of the cell-donor mouse and thus represent a genetically fully compatible 

 recipient, similarly to an ordinary F 1 hybrid. The use of such hosts for critical trans- 

 plantation tests would permit cleaner experimentation with old lines in tissue culture 

 than now practiced, and all sources of erroneous interpretation now stemming from the 

 homograft reaction would be eliminated. 



Only large-scale studies of the most rigorous design will illuminate the many 

 important problems that await answer regarding the probability of the malignant 

 change in vitro, its relationship to intrinsic cellular and host factors, and to extraneous 

 influences such as carcinogenic agents, radiation, viruses, and the cultural environment. 

 A clear-cut test system is also essential to follow the change from normal to malignant 



