356 IMMUNOGENETICS 



and of genes, in terms of the antigens they control. This point of principle will not be 

 discussed further here; it has been considered in detail elsewhere. 981 - 985> 1286 



As to the methodology of absorptions themselves, the details will be found to 

 depend to some degree on the particular system to be investigated. In many systems, 

 such as rabbit antisera to cattle blood, or an antiserum produced by one cow against 

 red cells of another, one simply mixes the packed, washed, absorbing cells with a 

 dilution of the antiserum to be absorbed in nearly equal quantities, allows the well- 

 mixed mixture to stand for a few minutes, then spins the cells down and repeats this 

 procedure on the supernatant with fresh absorbing cells until all of the antibody 

 reactive with the absorbing cells has been removed. This process may be complete in 

 two successive absorptions (as with most cow-anti-cow sera), or may require four or five 

 successive absorptions (as with rabbit-anti-cow serum). The objective here is get out 

 all of the antibody with which the absorbing cells can react; further absorptions with 

 these cells or with negative cells then have essentially no effect on the reagent. In 

 many other systems, however, the amount of packed red cells used in an absorption, 

 and the number of absorptions, must be cautiously controlled; a procedure described 

 by Race and Sanger 1036 as that in use at the Medical Research Council for the prepara- 

 tion of anti-human M and N sera is a useful reference. 



The reason generally given for the necessity of proceeding with caution in situations 

 like that for anti-human M and N is that the anti-M and anti-iV antibodies are not 

 fully specific for the corresponding antigen, but will in fact unite with low avidity to 

 the other antigen as well. Even the relatively rare cases in which human beings 

 develop anti-A 7 prove to be specific for N at only certain temperatures; they react with 

 M cells at lower temperatures, or after enzymatic treatment of the cells. 585 The plant 

 anti-A 7 can be adsorbed by M cells, although it is much more easily dissociated from 

 them than it is from N cells. 784 Similar situations are encountered with reagents for 

 red-cell differences among mammals other than man. They lead, of course, to 

 difficulties; for example, in such a situation a questionable test result can be checked 

 only with difficulty by an absorption test because even a negative cell may remove or 

 reduce the antibody in question. They also make it difficult to judge the specificities 

 of cells by tests on eluted antibody after absorption. Techniques for elution of antibody 

 will not be discussed in detail in this chapter ; one of the easiest involves the principle 

 of adsorbing antibody to the surface of the cell at a low temperature, washing the cells 

 gently in the cold, and then warming and shaking the system in a saline medium. 

 The antibody may then elute from the cell surface into the saline, from which the cells 

 can be removed by quick centrifugation at the higher temperature. 



Another basis for deviations from ideal behavior in absorptions has been sug- 

 gested by the work of Jacquot-Armand et a/. 659 It appears that the red cells of certain 

 animals will remove a fraction of human anti-5 only when the cells are used in large 

 excess, and that this adsorption is promoted by nonspecific factors in serum, such as 

 added serum from persons of type AB, which of course lacks the anti-i? antibody itself. 

 More work should probably be done along this line ; if it is true that nonspecific co- 



