GENETICS OF SOMATIC CELLS 413 



centromere and that marker as fractions of the total. This mapping, which led to the 

 identification of linkage groups, the order of the genes and their location, turned out to 

 be fully dependable; indeed, from the qualitative point of view it was more dependable 

 than meiotic mapping, due to the rarity of multiple exchanges in s.c.o. and to the 

 occurrence of haploidization without crossing over. 



It has been repeatedly emphasized by Pontecorvo and his associates 1011, 1012 that 

 there was no reason why s.c.o. could not be used, if it occurred, for establishing the 

 formal genetics of somatic cells in cultures of tissues and organs from higher organisms. 

 Even processes akin to haploidization would be of great value. In Aspergillus, most 

 haploids originate by nondisjunction as aneuploids, which are rapidly reduced to the 

 monosomic condition by accidental selection in favor of the balanced haploid. No 

 similar process is known to occur in somatic cells of higher organisms; although 

 aneuploidy is common in established strains of cells in vitro, it usually leads to an increase 

 rather than a decrease of the chromosome number and no truly or nearly haploid cells 

 have been propagated in culture so far. 



For these reasons, it appears to be of considerable importance to find out whether 

 s.c.o. occurs in mammalian cells or not. Two phenomena have been described that 

 might be possibly interpreted as such. One is the somatic variation of human ery- 

 throcytic antigens, the other the variation of //-2-isoantigens in heterozygous tumors. 

 The former has been studied recently by Goudie 461 and by Atwood and Scheinberg. 37 

 The latter authors found that exceptional red cells lacking A agglutinogen were present 

 in the peripheral blood of normal A or AB persons, comprising about 0. 1 per cent 

 of the erythrocytes in heterozygous young adults. Their phenotype agreed in every 

 way with the expected consequences of the loss of the A allele. Some cells apparently 

 lacking B agglutinogen were also found in AB blood and they were phenotypically A. 

 Nonspecific inagglutinability was considered to have been ruled out by the fact that the 

 exceptional cells behaved in the same way as unchanged cells when typed for MN 

 components. 



Considering the question whether these cells represent changes at the genetic level 

 or phenocopies, Atwood and Scheinberg 37 suggest that this might be judged from 

 studies on the distribution of exceptional cells among different individuals, their 

 relation to the age of the individual, to heterozygosity, and to mutagenic agents. 

 Evidence on these points is mostly unavailable at present, although a very recent study 

 of Scheinberg and Reckel 1166 strengthened the genetic explanation by showing that the 

 frequency of cells with the inagglutinable phenotype can be increased by radiation. 

 Nevertheless, the argument remains rather hypothetical and it cannot be excluded 

 that the exceptional cells represent phenocopies rather than true genetic changes, but, 

 if so, they are at least accurate facsimiles of genuine mutants. While Goudie 461 

 considered mitotic crossing over as the most probable explanation, Atwood and Schein- 

 berg believe that mutation is more likely, since loss of A 1 was accompanied by a great 

 increase in the //-substance which they would not have expected in B homozygotes, 

 such as would arise by crossing over. This argument is highly conjectural, however, 



