424 GENETICS OF SOMATIC CELLS 



cells. The number of such correlated studies is quite limited as yet, however. 



The following sections will deal with a more detailed consideration of the various 

 experimental systems in which phenotypic marker characteristics have been used for 

 the study of variation in populations of neoplastic cells in vivo. 



Isoantigens and transplantation characteristics. — Antigens are believed to reflect the 

 specificity of genes more directly than other cellular characteristics. 1435 The genetic 

 mechanism of determination for several isoantigenic systems is known, at least in the 

 mouse, and this makes them particularly suitable as markers for studies on variation in 

 populations of tumor cells. Owing to the impressive developments in transplantation 

 genetics during the last decades, it is now well established that the transplantability 

 of normal and tumor tissues depends on a fairly large number of genes, called histo- 

 compatibility genes, located at different chromosomal loci and subject to considerable 

 genetic variation in all species of vertebrates in which they have been studied. In the 

 mouse, one particular locus, called H-2, has been studied most extensively, 1249 and its 

 allelic identity in the donor and the actual recipient appears to be one of the most 

 important prerequisites for transplantability (for transplanted tumors, donor means 

 the animal in which the tumor has originated). For this reason, H-2 is often referred 

 to as a strong locus. It is also well known that the various allelic forms of H-2 

 determine the specificity of a whole complex of cellular components that can act as 

 isoantigens in other animals devoid of the same components and are then demonstrable 

 by serologic techniques. 18, 456, 587 The number of identified components determined 

 by a given H-2 allele is continuously growing, and probable crossovers within the H-2 

 system indicate that this is actually a compound system of pseudoalleles. 985 



Histocompatibility characteristics have been studied in tumors ever since the earliest 

 days of transplantation experiments on inbred mice. Early results of Tyzzer 1334 and 

 of Little and Tyzzer 802 were interpreted by Little 791, 795 as indicating that the take 

 and growth of a transplanted tumor in a new host depended on the simultaneous 

 presence in the recipient of more than one gene, derived from the donor of the original 

 tumor. Strong and Little 1299 have shown some years later that two tumors of closely 

 similar origin and histology grew in distinctly different frequencies when inoculated 

 simultaneously on opposite sides of the same animals belonging to a segregating hybrid 

 generation and derived from an outcross between the strain of origin of the tumor 

 and an unrelated strain. For a given tumor, this percentage of takes (later often 

 referred to as "gene requirement") could change in the course of time. Strong 1295 

 has isolated two rapidly growing variant sublines from an adenocarcinoma of the DBA 

 strain which differed from the original tumor with regard to their gene requirements. 

 While the percentage of takes in the original line of a segregating DBA x A F 2 hybrid 

 population were compatible with a requirement for six genes derived from the DBA 

 strain, one subline gave a two-factor and the other a one-factor ratio. This was 

 interpreted to mean that a genetic change has occurred in the variant lines. Similar 

 changes, always proceeding in the direction of decreased specificity, have been des- 

 cribed by Bittner 108 and by Cloudman. 208 Since the changes appeared to be sudden 



