460 GENETICS OF SOMATIC CELLS 



and selective experimental systems are now available for the large-scale exploration of 

 these possibilities. 



Meanwhile, existing information is often conjectural but by no means uninteresting. 

 Somatic crossing over occurs in Drosophila, and the genetic and environmental factors 

 influencing its frequency have been throroughly analyzed. In Aspergillus, somatic 

 crossing over has permitted genetic mapping in no way inferior to meiotic mapping, 

 and it has been pointed out that this mechanism could be used for the genetic analysis 

 of somatic cells even in the absence of genetic transfer, provided that it occurs in higher 

 organisms. Evidence on this point is not conclusive, although recent findings on the 

 variation of erythrocytic antigens in man and tissue isoantigens in tumors derived 

 from heterozygous, F 1 , hybrid mice can be interpreted as possible mitotic crossovers. 

 More information is urgently needed on this point. 



Somatic mosaicism due to chimerism, mitotic nondisjunction, and somatic muta- 

 tions is receiving increasing attention. Rare cases involving the gonads in addition to 

 somatic tissues have been very helpful in proving the genetic nature of the underlying 

 changes. Sometimes somatic mutations reveal new information about cellular lineage 

 in development. They can be induced by X rays like germinal mutations and suitable 

 experimental designs permit the estimation of their frequency, being apparently of 

 the same order as of comparable germinal mutations. In maize and in Drosophila, 

 highly informative studies are available about the genetic control of somatic variega- 

 tion in various tissues. This phenomenon is so highly regulated and orderly that 

 attempts have been made to bridge the gap between genetics and development by 

 constructing theories of differentiation based on controlling elements at the chromoso- 

 mal level, such as dissociators, activators, and modulators. Whatever the future of 

 this theory, it is obvious that somatic variegation and its genetic control are of the greatest 

 importance for the field of somatic-cell genetics. Changes at the genetic level have 

 been also implicated as a basis of the clonal theory of antibody formation; this is as 

 yet entirely at the speculative level, however. 



Neoplastic cells in vivo are valuable tools for the study of somatic variation, although 

 they represent a special case of somatic cells and the possible relevance of the findings 

 for comparable normal cells will have to be established from case to case. Among the 

 phenotypic marker characteristics that can be used for studies on population dynamics 

 in vivo, isoantigens of the histocompatibility system, drug resistance, hormone de- 

 pendence, ability to grow in the dissociated free-cell form of an ascites tumor and 

 correlated surface characteristics, and, in some special cases, tumor-specific antigens 

 and other cellular products may be mentioned. Among those tested in this respect, 

 the isoantigenic system has the highest selectivity and represents the only marker with 

 a known genetic determination mechanism. Isoantigenic variation can be studied 

 particularly well in heterozygous tumors of Fj-hybrid origin, since variant sublines 

 can be selected, compatible with one or the other of the parental strains and charac- 

 terized by the loss of specific //-2-determined isoantigens. Such variants could be 

 selected in one step or through several intermediate steps ; different types of variants 



