248 THE ANTIGEN-ANTIBODY REACTIONS 



eluded that by heat treatment in an acid solution, diphtheria antitoxin globulin was 

 partially denatured, and could be refined by proteolytic enzyme hydrolysis, leaving an 

 undenatured portion which carried most of the antitoxic activity. Similarly, Petermann 

 (1942) demonstrated that diphtheria antitoxin was spUt into approximate halves of 

 different solubihties by papain digestion, one half carrying the antitoxic activity, the 

 other being inert. If digestion was prolonged so that the halves spUt into quarters, no 

 antitoxic activity remained. The process appeared to be one of enzyme cleavage, and 

 not a denaturation. The antibody loses most of its original antigenic activity as a result 

 of its " proteolytic refinement " (Weil, Parfentjev and Bowman 1938). The change is 

 reflected in the distribution of electrophoretic components. Refined antisera show a 

 diminution or absence of albumin, disappearance of the T component in diphtheria anti- 

 toxin and its replacement by a new y component (van der Scheer, WyckoflF and Clarke 

 1941c), or a relative increase in the /3 and )' components (Fell, Stern and Coghill 1940). 



From this fact, and the varying distribution of antibody among the components, 

 it is clear that the characterization of an antibody globulin by its electrophoretic mobility 

 has no necessary connection with the fact that it carries immunologically active groups. 



An interesting example of partial digestion of antibody is recorded by Kalmanson 

 and Bronfenbrenner (1943). A bacteriophage just neutralized by homologous rabbit 

 antibody was restored to activity by papain digestion of the globulin on its surface. The 

 digestion of over-neutralized bacteriophage, on which there were many more antibody 

 molecules, did not reverse the neutralization. Some part of the antibody apparently 

 remained on the phage particles after digestion, for just-neutralized phage after digestion 

 would sensitize guinea-pigs to rabbit globulin ; and over-neutralized j^hage after digestion 

 was found to be incapable of stimulating phage antibody on injection into a rabbit, pre- 

 sumably because the antibody remnants formed an obscuring layer on its surface. 



Rosenheim (1937) has drawn attention to an increase in the resistance of antibodies 

 to pepsin and trypsin hydrolysis that takes place as immunization proceeds. She found 

 that " H " bacterial agglutinins (see p. 276) from late bleedings of an immunized rabbit 

 were resistant to a degree of digestion that destroyed 80 per cent, or more of agglutinins 

 from the first bleeding. There did not, however, appear to be a corresponding increase 

 in the resistance of the serum proteins to digestion. " " antibodies did not become 

 resistant in this way. Both Pappenheimer and Robinson (1937) and Pope ,(1939a) noted 

 that the resistance of diphtheria antitoxin to proteolytic refinement varied with the stage 

 and the method of immunization of the horse. 



Unity and Diversity of Antibodies. 



Immunization with a single antigen yields an antiserum which can display 

 a variety of specific phenomena ; for instance, precipitation, neutralization of 

 toxin, protection of an infected animal. In the interpretation of antibody reactions, 

 there are two questions about the variety of antibodies that need an answer. 

 In the first place, are all these manifestations due to a single kind of antibody, 

 or does a precipitin, for example, differ fundamentally from an antitoxin, when 

 the toxin serves as antigen in both reactions ? In the second place, in a solution 

 that displays a given kind of serological activity, are all the molecules of antibody 

 similar ? In other words, in a monospecific antibody solution, are there antibodies 

 with different functions, and are those of the same function homogeneous ? 



The view that precipitins, opsonins, antitoxins, etc., differed in kind, held the 

 field for many years ; but the unitarian hypothesis that these were due to the 

 same kind of antibody in different circumstances has now gained almost universal 

 acceptance. This has been due mainly to the work of Dean (1917) in this country, 

 to that of Zinsser (1921) in America, and to Nicolle and Cesari (1922) in France ; 

 though a similar view had been adumbrated by Bail and Hoke (1908). We have 

 seen how Dean established the essential identity of precipitation and complement 



