GENETICS OF SOMATIC CELLS 423 



The choice of suitable phenotypic marker characteristics for the study of somatic 

 variation is influenced by several considerations. It is desirable to have a fair knowl- 

 edge of their genetic determination mechanism. This should be sufficiently variable 

 in nature to permit the identification of a series of alternative forms. It is essential, 

 furthermore, to have some kind of a selective device that permits the concentration of 

 rare alternative forms (variants) from large populations of cells. 



Among the phenotypic characteristics encountered in populations of neoplastic 

 cells in vivo, some are common to most or all normal neoplastic cells of the organism 

 while others are more or less distinctive for the neoplastic transformation itself. The 

 characters in the former category studied best are represented by the isoantigens of the 

 histocompatibility system in the mouse. These have been explored in considerable 

 detail by several schools of transplantation geneticists, immunologists, and general 

 biologists among whom Little, 794 Strong, 1292 Snell, 1248 Gorer, 456 Hoecker, 587 Amos, 18 

 Medawar, 862 and their co-workers may be mentioned particularly. It is neither neces- 

 sary nor feasible to discuss these developments here; several excellent review articles 

 of recent date are available on the subject. 129 - 456 - 528 - 529 - 862 - 1248 - 1249 Since these 

 characteristics are common to all cells of the organism, they can be subjected to straight- 

 forward genetic analysis by crosses, and their genetic determination mechanism is 

 fairly well known, at least in the mouse ; other species are being studied to an increasing 

 extent 62, 348, 575- 576- 675 



The second category of properties would include the unit characters of tumor 

 progression. 399, 401 Their mechanism of determination is unknown, but nevertheless 

 they are of great interest because of their direct relationship to the phenomena of malig- 

 nancy. Drug resistance, the ability to grow in the ascites form, and, to some extent, 

 invasiveness and ability to metastasize are such characters that have already been 

 studied in experiments designed to investigate the population dynamics of progression. 



The selectivity of the experimental systems involved has not been studied to a very 

 considerable extent; some reconstruction experiments with artificial mixtures of differ- 

 ent types of cells are available for the histocompatibility system, 723 for some cases of 

 drug resistance in murine leukemia, 720 - 725 - 1018 and for the convertibility of a certain 

 solid murine sarcoma into the ascites form. 721 Among these, only the first permits the 

 design of absolutely selective systems in which the selective force utilized (the homo- 

 graft reaction) is able to inhibit the growth of one cellular phenotype completely while 

 leaving an alternative form undamaged. In the other two systems, growth inhibition 

 of one type is not complete, and the selective advantage of deviating variants is rela- 

 tive rather than absolute. This makes them less selective; in the particular cases 

 investigated by reconstruction experiments, the minimum frequency of demonstrable 

 variants has been of the order of 10" 6 in the case of amethopterin resistance 720 and 

 4 x 10 ~ 5 with the ascites system. 726 



Studies on chromosomal morphology may yield direct information about changes 

 at the genetic level. Correlated studies on biological changes and alterations of 

 chromosomal morphology may lead to a true cytogenetics of populations of neoplastic 



