254 THE ANTIGEN-ANTIBODY REACTIONS 



The results obtained with azo-proteins were criticized by Glutton, Harington 

 and Mead (1937) on the ground that the diazo link is highly artificial and is never 

 found in natural substances. They devised means of coupling carbohydrate to 

 tyrosine, by an 0-glucosido linkage, and attaching the compound to the NHg 

 groups of protein. Subsequent work (Glutton, Harington and Yuill 1938), how- 

 ever, showed that the nature of the linkage to the protein had no observable effect 

 on the modification of antigenic specificity induced by the glucosido-tyrosyl groups 

 themselves. Another objection to the diazo method is the drastic chemical treat- 

 ment to which the proteins are subjected during diazotization. Hopkins and 

 Wormall (1933a, b) were able to modify the specificity of proteins by the much 

 gentler action of phenyl iso-cyanate in alkaline solution. The iso-cyanate reacts 

 with free — NHg groups to form a substituted urea. The results of chemical 



A3O5U2 



ATQXYL 

 |p-amino- benzene arsinic acid] 



N=N-CL 



+ TYJ^PSIME. 

 CH2.CH-NH2-COOH. 



I 



H2O5A5 < > N = N IJ N = N < >&sO^«? 



OH. 



Fig. 41. 



modification by all these methods are essentially similar, and we may take those 

 obtained by the extensively employed azo-proteins as valid. 



In general the protein to which such a grouping has been attached will retain 

 its species specificity, though usually with some loss of potency. Thus an antigen 

 prepared by coupling atoxyl to horse globulin will react with an antiserum prepared 

 against any atoxyl-azo-protein, in virtue of the atoxyl grouping ; and with an anti- 

 serum prepared against horse globulin or against an azo-protein prepared from 

 horse globulin, in virtue of the specific horse-globulin groupings. This difficulty 

 is overcome by coupling any grouping that it is desired to study to two immunologi- 

 cally unrelated proteins. One of these is used as the antigen for the preparation 

 of the antiserum, the other is used for the in vitro tests. Thus, atoxyl may be 

 coupled with horse globulin, and the atoxyl-azo-protein so prepared may be used 

 for the immunization of a rabbit. The serum so prepared will react with this antigen 



