374 CELL HEREDITY 



rlu'ic is cNcii tlu' possibility that gfiictic oxchangc may occur l)ctwcen 

 the timiDi cells and those of the host, for the histocompatibihty antigens 

 are toiinil in nianv, if not all, of the cells of the body. There seems to 

 be no tissui' specificitv in the content of these substances. But, of 

 course, w hen we ha\e a mosaic, we expect differences as a result of varia- 

 tion not restricted to a given tissue. Such cases are known. For 

 example, the ABO antigens detected on the red blood cell surface are 

 usually also secreted in the sali\a, but individuals are known in which 

 the antigens in the two sites are different. Furthermore, it appears that 

 everyone is mosaic to a small extent for the antigens of the red blood 

 cells. Although these cells arc mixed by the circulation, the blood- 

 forming sites in the bone marrow have small patches of somatic variants. 



The demonstration of this fact requires the development of sensitive 

 techniques for the detection of very small frequencies of red cells of 

 aberrant type. One method entails the labeling of a sample of blood 

 cells with radioactive chromate, Na2Cr^^04, followed by treatment with 

 a specific agglutinin. About 15 per cent of the cells, and hence of the 

 radioactivity, remains in the supernatant. Unlabeled cells of the same 

 type are then added, and agglutination is carried out once more with 

 the result that the radioactivity of the new supernatant is further re- 

 duced. This procedure is repeated until the radioactivity in the super- 

 natant reaches a stable level. A knowledge of the dilutions allows cal- 

 culation of the proportion of nonagglutinable cells. It can be seen from 

 Figure 12.10 that each agglutination step removed the same fraction of 

 marked agglutinable type-A cells and that eventually there was a residue 

 of cells that would not agglutinate. These nonagglutinable cells can be 

 independently isolated and prove to lack the Aj antigen. 



But two types of nonagglutinable cells were found to be present by 

 the use of different specific agglutinins. Type A, blood from which 

 the type Aj cells were removed by A, agglutinin showed a further 

 decrease in the proportion of nonagglutinable cells upon the addition of 

 A.) agglutinin (Figure 12.10). Thus it was shown that type-A, blood 

 had not only about 10~'^ type O cells but about the same frequency of 

 type A^ as well. Similarly, A^B blood had both B and A.B cells in low 

 frequency. The () and B cells would be expected to result from the loss 

 of the gene A^, but the A^ and A<,B cells may inxolve a replacement of 

 gene A, by A.,. Since neither A, nor A., is ever lound in cis arrange- 

 ment with B. this result could not be obtained bv somatic crossing 

 over with an ostensible A,A2 cistron. Furthermore, in different blood 

 samples, there is no correlation between the frequency of cells with A2 

 antigen and with neither ,\, nor A., antigen. Therefore the appearance 

 of the \aricUit cells cannot be explained b\' anv process leading to the 



