502 PROTOPLASM 



theoretical grounds, it is possible that lipoids exhibit specificity. 

 Carbohydrates had not, at the time when Wells wrote (1929), 

 been found to be antigenic, though he recognized the fact that 

 they are capable of exhibiting immunological specificity. It is 

 now known that certain polysaccharides are, and that lipoids 

 may be, antigens. 



The evidence in favor of a protein nature of antibodies is 

 reasonably convincing. Precipitation, cataphoresis (isoelectric 

 point), stability, and destruction (with heat, alcohol, etc.) all 

 show a close correspondence between antibodies and proteins. 

 The antibody protein is presumably a globulin fraction. Experi- 

 ments by S. Mudd on cataphoresis (bearing on the problem of 

 phagocytosis) are reasonable evidence of the globulin nature of 

 antibodies. 



Antigens and antibodies being recognized as (mostly) protein, 

 the next problem has to do with the mechanism by means of 

 which the former are rendered nontoxic by the latter. The 

 question is a basic one in protein specificity and first received the 

 serious attention of the German chemist Paul Ehrlich, to whom 

 much of the early development of immunology is due. He 

 represented the molecules of the antigen and the antibody as 

 geometrical figures which fit into each other like a key into a lock. 

 Restated in other terms, the hypothesis of Ehrlich implies that 

 immunological reactions between proteins are due to arrange- 

 ments of the constituent parts (amino acids) of the molecules. 

 The arrangement responsible may not be of the molecule as a 

 whole but of a determinant group in the antigen and a receptor site 

 in the antibody. (These terms are simply convenient expressions 

 for specific molecular orientations; chemically, they would be 

 called atomic groups, radicals, or side chains.) As Marrack puts 

 it, specificity depends upon a specific configuration of atoms 

 representable by chemical formulas. Such a hypothesis of 

 immunological reactions is a confirmation of Ehrlich's views. 

 Those who oppose the lock-and-key hypothesis regard the 

 combination between antigen and antibody as one of adsorp- 

 tion. But as adsorption bonds may apparently be anything 

 from primary valence to a loose electromagnetic attraction, it 

 would be difficult to distinguish between adsorption and an 

 interlocking of atomic structures. 



