SARA E. BRANHAM 719 



similar to those of the homologous antigen, especially if the diazo-group was similarly 

 located in the aromatic nucleus. Landsteiner later" repeated some of these experi- 

 ments, using the anaphylaxis reaction instead of the precipitin test. Guinea pigs were 

 sensitized through injections of one azoprotein (horse serum treated with diazotized 

 para-arsanilic acid) to other proteins containing the same azo-group in chicken serum. 

 Injections of the related simple diazotized components alone did not cause shock, 

 but apparently induced a state of anti-anaphylaxis or desensitization, since subse- 

 quent injections of the azoprotein were without marked effect. 



Another series of azoproteins was made^ by treating many quite unrelated pro- 

 teins with two diazotized compounds, metanilic and para-arsanilic acids. Immune 

 sera prepared with these were specific, precipitating all proteins which contained the 

 homologous azo-group, whether it were rabbit serum, casein, gliadin, or zein. Com- 

 pounds made with gelatin or peptones gave only negative results. Here species 

 specificity was apparently entirely lost, all specificity being carried by the intro- 

 duced groups, the protein part of the molecule serving only to incite antibody forma- 

 tion. 



These diazotized compounds are not in themselves antigenic, and when added to 

 immune serum prepared with proteins treated with the homologous azo-compound 

 they cause no precipitate to form. Nevertheless, they seem to unite specifically with 

 the antibodies present in the serum, for when the homologous azo-protein is added 

 later no reaction occurs. This inhibition of the precipitin reaction by the correspond- 

 ing azo-compounds is shown in an extensive series of experiments with organic acids.^ 

 It is probably analogous to the desensitization produced by injecting the compounds 

 alone into guinea pigs sensitized with azoproteins. These experiments led Land- 

 steiner to the conclusion that relatively small portions of the large antigen molecules 

 are responsible for their specificity. 



Toxins, too, have been changed by chemical means. Ramon^ has produced "ana- 

 toxins" by treating diphtheria toxin with formaldehyde. The action of the toxic 

 radical in the molecule is completely destroyed. The antigenic properties are, how- 

 ever, unaltered, either in activity or specificity. 



SUMMARY 



Comparatively little is actually known about the chemistry of antigens. Nearly 

 all whole native proteins are more or less antigenic. It appears that astonishingly 

 minute quantities of some of these are sufficient to stimulate antibody formation. 

 That an amount of antigen containing as Httle as one- to four-millionths of a gram 

 of protein can be entirely responsible for such a biological reaction is surprising. Such 

 findings lead some to question whether or not substances other than protein present 

 in an antigen may have some influence on the production of antibodies. It has not 

 been definitely proved that any other kind of substance, in a pure state, can be truly 

 antigenic. Nevertheless, there is much evidence which indicates that some lipins and 



'Landsteiner, K.: /. Exper. Med., 39, 631. 1924. 



^Landsteiner, K.: Biochem. Ztschr., 93, 106. 1919. 



^Landsteiner, K.: ibid., 104, 280. 1920. 



''Ramon, G.: Ann. de I'lnst. Pasteur, 39, i. 1925. 



