440 GENETICS OF SOMATIC CELLS 



will have to be obtained, as in microbial genetics, by replica plating or some other 

 method of indirect selection. 186, 776 Another approach concerned with the relation- 

 ship of somatic-cell variation to drug resistance is the study of the karyotype. It has 

 been observed repeatedly 89, 527 that the appearance of drug-resistant lines may be 

 accompanied by changes in modal number of chromosomes or with regard to visible 

 chromosomal markers, but this was not a regular finding and it was not clear whether 

 the variation exceeded what could have been expected in a series of unselected, random- 

 ly isolated clones. In the study of Biesele el a/., 89 all five amethopterin-resistant sub- 

 lines of leukemia L1210 lacked a large submetacentric chromosome that was present 

 in the modal cellular type of the parental line and four other amethopterin-sensitive 

 sublines. It was pointed out that the relation between resistance and the loss of the 

 marker had a low probability of being fortuitous. 



Recently, Vogt 1349 published some interesting studies particularly stressing the 

 problem of whether the properties of drug resistance for the various lines of cells are 

 causally related to the differences in their karyotypes or whether they are only an 

 additional expression of the variability between tumor cells. She studied two clonal 

 lines of HeLa cells in vitro and compared them to variant sublines adapted to growth on 

 suboptimal media or in the presence of the drug, aminopterin. Each karyotype was 

 classified by the total number of its chromosomes and by the number of large chromo- 

 somes corresponding to the three largest pairs in the normal idiogram. Two sublines 

 adapted to growth under suboptimal nutritional conditions and three sublines showing 

 an increased resistance to aminopterin all had karyotypes which had not been observed 

 in the parental lines. Also, clones selected for variant cellular or colonial appearance 

 differed in their karyotypes from the parental lines and one another. Special 

 significance was attributed to the finding that lines selected and maintained under strong 

 selection pressure also maintained a stable karyotype. This was taken to indicate that 

 the selective value of a cell depends on its karyotype. This could be best explained 

 on the assumption that the karyotype determined the phenotype. 



While these findings support the assumption, so convincingly proved with various 

 microorganisms, that the variation underlying the development of resistance to drugs 

 in tumors is probably localized at the genetic (that is, DNA) level, they do not prove it. 

 It may be well at this point to place emphasis on the truism that tumor cells are not 

 microorganisms. Neither the stability of the changes nor the apparently random 

 appearance and clonal growth of the resistant variants does critically prove that changes 

 at the genetic level are responsible. When dealing with cells of higher organisms, 

 even if neoplastic, it must be realized that they may be subject to developmental 

 processes which, although still among the most obscure in biology, are regarded by the 

 majority of workers as not being akin to mutations at the genetic level. They are 

 usually regarded as epigenetic changes par excellence, caused by modifications of genie 

 expression rather than by a change of the genetic information itself. A progressive 

 limitation of genie expression may occur in the course of differentiation ; this "'may 

 be of such a natui'e that in each cell only a special complement of loci maintains its 



