4-32 GENETICS OF SOMATIC CELLS 



mixed spleen, liver, and kidney tissue, given intraperitoneally exactly 7 days prior to 

 tumor inoculation. According to Kaliss, 674 when enhanced tumors are tested in 

 preimmunized mice, they usually fail to grow, at least when the time between the 

 immunizing (first) inoculum and the grafting of the enhanced tumor is short enough 

 (for example, 7 days) . The results showed that the variants grew equally well in both 

 types of preimmunized mice. 



Thus, variant formation appears to be different from immunologic enhancement. 

 Although some mechanisms can be eliminated in this way, it is still very difficult to 

 identify the mechanisms really at work. The ambiguities are due to the impossibility 

 of distinguishing between genotype and phenotype. The irreversibility of the variants 

 and the fact that they could be obtained from heterozygous but not from homozygous 

 tumors are suggestive of a genetic phenomenon. However, stable changes of antigenic 

 phenotype have been described at the cellular level, for example, in Paramecia and in 

 bacteria in the absence of any genotypic change. 1257 Such changes, assumed to depend 

 on "mechanisms that regulate the expression of genetic potentialities," as contrasted 

 to "truly genetic mechanisms that regulate the maintenance of the structural informa- 

 tion" 325 have been called "epigenetic" by Nanney. 934 Unless ordinary genetic 

 tests can be performed, it is not possible to distinguish conclusively between genetic 

 and epigenetic mechanisms and only indirect arguments can be applied. The only 

 guide is that epigenetic changes are, as a rule, less stable and can be induced in a 

 predictable way. 934 



The epigenetic explanation is rendered less likely by the high stability of the parent- 

 compatible variants upon F x passage and the impossibility of reestablishing the missing 

 serologic phenotype in the course of this passage, or by antiserum treatment in vitro 

 or by forcing the tumor through the opposite, incompatible parental strain by using 

 newborn mice as recipients. 716 Furthermore, in some cases, the variants were demon- 

 strated to have been preexistent in the original Fj-hybrid, tumor population by 

 chromosomal studies 64 and by isolating variants directly in preimmunized parental hosts 

 for which the background growth of the original unchanged cell type was reduced 

 to a minimum. 545, 716 The preexistence of the variants in the original population, 

 before it was exposed to the selective environmental conditions, makes an epigenetic 

 mechanism less likely, since the latter is usually triggered by altered environmental 

 conditions. 



Turning to genetic mechanisms, point mutation is not very likely since several 

 antigenic specificities were often lost simultaneously, 64 - 545, 717 some of which are 

 believed to be controlled by at least two different chromosomal sites (but see Owen 985 

 for a different interpretation). Also, all changes so far identified were due to losses 

 rather than to new, alternative specificities which would be equally probable if point 

 mutations were responsible. Mutagenic treatment with X rays and triethylene- 

 melamine did not give consistent results, 265, 716, 717 although increased variant 

 formation was evident with some tumors that already exhibited a spontaneous tendency 

 to throw off variants of the type in question. 38, 265, 717 



