GENETICS OF SOMATIC CELLS 433 



Gross chromosomal changes of various kinds are conceivable as an explanation 

 and within the realm of experimental approach. One such study has been made by 

 comparing an heterozygous sarcoma with its 10 different variant sublines, all isolated 

 in and compatible with the same parental strain. 64 With the exception of two variants, 

 derived from the same original inoculum pool and that showed closely identical karyo- 

 types and probably represented the repeated isolation of the same clone, all other 

 variants differed from each other and from the original tumor with regard to their 

 chromosomal complement. It is conceivable that chromosomal changes of various 

 kinds have occurred in these lines, involving the chromosomal segment carrying H-1. 

 Alternatively, the change may have a more indirect chromosomal mechanism, due to 

 the remodeling of the genetic background that may lead to changes of modifiers, 

 suppressors, position effects, and the like. It cannot be excluded, however, that the 

 chromosomal differences observed merely reflect clonal variation in the tumor without 

 being causally connected with variant formation. This suspicion is strengthened by 

 the fact that methylcholanthrene-induced sarcomas, including the MSWB tumor used 

 in this study, do have a wide range of chromosomal variability to begin with. Lympho- 

 mas are much less variable, and, in an analogous study, Hellstrom 545 could not 

 demonstrate any gross chromosomal differences between ¥ 1 lymphomas and their 

 derived variant sublines. 



Mitotic crossing over, discussed in detail in the section on normal cells, remains 

 perhaps the most probable explanation. It would lead to apparent antigentic losses 

 and no gain of new specificities. It would yield homozygous, rather than hemizygous 

 variants. In this connection, it is of interest that in his work with lymphomas of 

 Fi-hybrid origin, Hellstrom 545 showed that variants selected for compatibility with one 

 parental strain (P x ) became similar to homozygous cells (of P x origin) with regard 

 to their sensitivity to the cytotoxic action of anti-//-2 isoantibodies (of the type P 2 

 anti-P l5 P 2 being the opposite parental strain entering the cross), and distinctly different 

 from the heterozygous cells of the original F 1 lymphoma from which they had been 

 derived, the F 1 cells showing a quantitatively lesser sensitivity. Also, as mentioned 

 previously, the P x variants grew more specifically and were less liable to take in the 

 opposite parental strain (P 2 ) than the original F 1 tumor, behavior similar to homozygous 

 tumors. Analogous results were obtained for a number of F : carcinomas and sarcomas 

 and their variants. 716 This, together with the total failure of persistent attempts to 

 switch established variants, compatible with one parental strain, to the opposite parental 

 type, again is in agreement with homozygous, rather than hemizygous constitution of 

 the variant cells (unless the cell losing both H-2 factors becomes inviable and thus 

 escapes detection). Thus, on the whole, there is a phenotypic similarity between the 

 variants and cells of homozygous origin. Unfortunately, such phenotypic similarity 

 cannot be taken as evidence for homozygous genotype. A genetically hemizygous 

 cell may mimic the homozygous phenotype, for example, if the products of the remain- 

 ing allele are capable of structurally replacing those of the missing allele. Therefore it is 

 impossible to decide between a phenocopy of the homozygote and truly homozygous cells. 



