452 GENETICS OF SOMATIC CELLS 



The findings were quite in accordance with the expectations. The aminopterin- 

 induced thymine deficiency was lethal for HeLa cells which ceased to proliferate but 

 increased in size and doubled their average content of protein. The increase in protein 

 could be prevented by the omission of a single essential amino acid, such as arginine, 

 from the medium. This resulted in a 10 5 -fold greater survival after 6 days when the 

 amino acid was supplied and aminopterin was removed as compared to survival on 

 complete medium. Ordinary HeLa cells which do not require glutamine showed a very 

 low survival in the presence of aminopterin in glutamine-free medium. In contrast, 

 glutamine-requiring auxotrophic mutants survived well and could be selected specifically 

 and efficiently, even if present in low frequencies. Small clones of auxotrophs could 

 be detected microscopically at an early stage (48 hrs.) after exposure to aminopterin, 

 due to their unswollen appearance, in contrast to the swollen look of the nonauxotrophic 

 cells. This technique appears very promising for the selection of clear-cut nutritional 

 variants. 



Drug-resistance markers have also recently entered the tissue-culture field. Some 

 of these studies have already been discussed in the previous section, particularly the 

 modified fluctuation test of Szybalski, 1307 suitable for the estimation of mutation rates, 

 and the experiments of Vogt, 1349 designed to establish the relationship between karyo- 

 type and cellular phenotype. Viral resistance also appears useful as a marker and 

 Vogt 1348 has reported on the chromosomal constitution of HeLa-czW variants resistant 

 to poliomyelitis virus. Increased resistance to X rays and ultraviolet may also be 

 workable. 1058 



Chromosomal markers have frequently been applied to the study of genetic varia- 

 tion in tissue culture. Most of this work has been done on established strains of cells. 

 No detailed consideration will be given here to this extensive field which has been re- 

 viewed and interpreted recently by Levan. 781 



One characteristic, useful as a marker and of the greatest interest by itself, is the 

 fascinating but still ill-defined malignization of normal-cell strains in vitro. Unfortu- 

 nately, this subject is rather difficult to evaluate at present. One reason is the surprising 

 and disconcerting finding of Rothfels et al. 1076 After a careful study of chromosomal 

 markers, these authors came to the conclusion that many so-called "altered" strains 

 of cells originally normal represent contaminations with some of the most extensively 

 used malignant lines, such as L cells or HeLa cells. Similar findings have been reported 

 by other cytologists. 391, 605 Even in cases where contamination can be critically 

 excluded, other factors impede evaluation. Whether because of the "anthropocentric 

 glamor of using human tissues," 773 or for other reasons, a large part of this work has 

 been done with human material. It is notoriously difficult to determine the normalcy 

 or malignancy of a line of human cells, since malignancy is essentially a matter of host- 

 cell relationship, and the appropriate, genetically identical host is not available for 

 transplantation tests. The use of an animal the heterograft response of which must 

 first be inhibited by such means as cortisone or X rays, or both, is by no means compar- 

 able. Failure of a line to grow may be due to residual graft reaction and not to lack 



