H. GIDEON WELLS 705 



ACTIVITY OF ANTIGENS 



Whatever the antigenic function of a protein may depend on, there are marked 

 differences in activity between different proteins which have not yet been explained 

 by anything we know about their composition or structure. In general, the more 

 soluble the foreign protein in the body fluids the better antigen it seems to be, but 

 this is not always true. For example, serum albumin seems to be much less strongly 

 antigenic than the less soluble globulins of the same serum. Casein is less antigenic 

 than many of the other proteins commonly used for experimental immunization. 

 Some toxic proteins, however, possess a most remarkable degree of antigenic activity. 

 For example, ricin, the poisonous protein of the castor-oil bean, injected into rabbits 

 in doses less than one-one-thousandth of a milligram (o.oooooi gm.), which is the 

 minimum lethal dose, will lead to antibody formation. Florence Seibert' has found 

 that some bacteria manage to live in distilled water, and that 1 500-1800 cc. of such 

 water concentrated to 20 cc. and injected into a rabbit sufl&ced to produce an agglu- 

 tinating serum active in a dilution of 1-8,000, although it could have contained not 

 more than infinitesimal amounts of bacterial protein. 



The anaphylaxis reaction demonstrates that almost any protein may be antigenic 

 in most remarkably small amounts. Thus, crystallized egg albumin has been found 

 to sensitize guinea pigs with a single dose of 0.000001 gm. or even less, and similar 

 activity has been demonstrated with other proteins. 



When a very small amount of one antigen is injected together with a much larger 

 amount of another, the effect of the former may be obscured or lost (concurrence of 

 antigens).^ The pH of the antigenic solution may modify its activity.^ 



TOXINS AS ANTIGENS 



Toxins are also antigens of high potency, although their protein nature has not 

 been fully established. They exhibit colloidal properties, but they diffuse more readily 

 than proteins usually do, and investigators have reported the isolation of active 

 toxins that do not give reactions for proteins. However, as shown by the figures for 

 the size of the lethal dose of a known protein poison, ricin, and the minute amounts of 

 bacterial protein capable of serving as an active antigen, failure to secure chemical 

 tests for protein in toxin preparations that are antigenic and toxic does not prove the 

 absence of proteins. It must be admitted that toxins resemble proteins in many 

 essential respects, especially in being attacked by proteolytic enzymes. 



They do not owe their antigenic effect solely to their toxicity, since when detoxi- 

 cated with formalin or some other agents they retain their antigenic activity,'' thus 

 resembling proteins which also are antigenic whether toxic or not. However, when 

 toxins are detoxicated with acids their antigenic power is lost, returning with reactiva- 

 tion of the toxin by neutralizing with alkali. In this process of inactivation and re- 



' Seibert, F.: Am. J. Physiol., 71, 621. 1925. 



- See Lewis, J. H.: /. Infect. D'ls., 17, 241. 1915; Imai, K.: Zlschr. f. Tmmiinitdtsforsch. it. e.xper- 

 Tlierap., 43, 312. 1925. 



■* Falk, I. S., and Powdermaker, F.: /. Infect. Dis., 37, 514. 1925. 

 ■• Ramon, G.: Ann. deVInsl. Pasteur, 39, i. 1925. 



