430 GENETICS OF SOMATIC CELLS 



In addition, K. E. Hellstrom 545 has studied the sensitivity of F^hybrid lymphomas 

 and their variants towards various isoimmune sera, by the direct technique of Gorer 

 and O'Gorman. 460 



All techniques gave essentially similar results. Variants which were able to grow 

 in a high frequency of preimmunized mice of their new parental genotype did not seem 

 to contain detectable amounts of isoantigens derived specifically from the oppposite 

 parental strain. Less clear-cut results were obtained with variants that grew in non- 

 immunized but not in preimmunized mice, but the opposite parental antigens were 

 still detectable in several cases by the sensitive method of provoking a second-set response 

 with preirradiated cells and by the direct cytotoxic test on the lymphoma cells. In 

 several cases, subsequent serial transplantation in nonimmunized mice of the parental 

 genotype led to an increased ability of such variant cells to grow in preimmunized 

 mice. When tested again at this stage, the specific antigens of the opposite parental 

 strain were no longer detectable with either technique. 



In conclusion, specific variants can be distinguished from false positives by their 

 selective compatibility with one of the parental strains, including the ability to grow 

 in hosts hyperimmunized against the opposite parental type, and the apparent loss of 

 //-2-determined isoantigens specifically derived from the opposite parent. Such vari- 

 ants have been found only with F x tumors so far and were limited to one of the parental 

 strains (to the exclusion of other coisogenic, resistant mice) carrying a foreign H-2 

 factor. False positives are less discriminative in their host requirements, grow in many 

 different genotypes to a varying extent, do not breed true to type on selective transfer, 

 and are completely or partially inhibited upon testing in preimmunized mice. They 

 occur with homozygous as well as with heterozygous tumors and are not limited to the 

 parental strains. The tendency to give rise to false positives varies considerably from 

 tumor to tumor. 



Further experiments showed that the specific variants were highly stable even if 

 returned to and carried in the F x hybrid genotype of origin in serial passage for pro- 

 longed periods of time. Their isoantigenic and transplantability pattern, whether 

 fully or only partially compatible with the parental host, could be reproduced with great 

 constancy even after prolonged F x passage. True variants were often more specific 

 in their transplantation behavior than the original F x tumor from which they had 

 been derived. This was particularly apparent with tumors capable of giving rise to 

 variants in both parental strains; the original F x tumor took in both parental types 

 with a certain frequency but variants selected in and compatible with Pi refused to 

 grow in P 2 hosts as a rule. 716 Attempts to switch variants compatible with one parental 

 type to the opposite parental type by passage through newborn mice of the opposite 

 parental strain were quite unsuccessful. 



Since four isogenic resistant lines characterized by four different alleles at the 

 H-2 locus were used in these studies, it was possible to study the question of whether a 

 tumor originating in a certain F 1 hybrid (for example, H-2 a H-2 s ) can give rise to 

 variants specifically compatible with a semi-isologous F 1 genotype, produced by 



