Normal antibodies in human sera that are capable of reacting with human B but not with O red blood cells are heterogeneous in type and degree of specificity. These antibody populations are fractionable by absorption with red blood cells from various animals. Sera from different group A individuals, for example, contain different relative activities of a β-antibody fraction that will cross-react with opossum red blood cells and one that will not cross-react.
In certain group O sera, most of the “β antibody” fraction that does not cross-react with opossum red blood cells turns out to be absorbable by (and thus cross-reactive with) human A cells. Inhibition with opossum saliva has closely comparable effects to absorption with opossum red blood cells. Other fractionations and tests have been performed using red blood cells from man, rabbits, guinea pigs, hamsters, cotton rats, Norway rats, chickens, sheep, and cattle, individually and in various combinations. In addition to considerable diversity in normal antibody specificities within particular sera, there are marked differences between sera from individuals of the same classical blood group, as well as between different blood groups.
Immune sera produced by injecting rabbits with human red blood cells also show antibody populations with diverse patterns of specificity and cross-reactivity.
Antibody fractions from different sources, cross-reactive with human A and B cells (but not with O), are not all the same; they vary in other details of their specificities and are often subfractionable by absorption with related antigens.
The experimental data on which the link between specific α and β antibodies has been hypothesized is reinterpreted in terms of the heterogeneous antibody populations demonstrated in this study.
Fractionation of sera by absorption with animal cells offers promise as a source of reagents for details of antigenic variation in human cells that are otherwise difficult or impossible to detect.
For example, reagents for a common pattern of reactivity to A, B, and certain rare "O" antigens can be obtained by adsorbing selected O sera with particular animal cells. However, the problem of removing all the α-specific antibodies, but leaving the cross-reactive antibodies, has not been solved.
Other implications of the observations reported in this paper, relevant to the origins and behavior of normal antibodies, have been discussed.