414 BACTERIOLOGICAL CHEMISTRY 



on the introduced determinant group. Thus iodo- and 

 bromo -proteins were found to have lost their protein 

 specificity and reacted only with antisera prepared against 

 themselves, although there was strong cross -reaction 

 between the iodo- and bromo -derivatives. The reactions 

 between both iodo- and bromo -antigens and antisera 



I 



were inhibited by di-iodotyrosine, H0<^ ^CH^.CH.COOH, 



which is apparently the corresponding hapten. Neither 

 tyrosine, H0<^ ^CHa-CH.COOH itself, nor di-iodophenol, 



NH2 

 I 



HO^ y, nor potassium iodide acted as haptens. 



I 

 Dibromotyrosine behaved as a hapten, but less strongly 

 than the iodo -derivative. 



Harrington has made use of tyrosine for coupling 

 various determinants to proteins. He prepared the 

 0-^-glucosidyl-tyrosyl-derivatives of gelatin and insulin 

 by condensing glucosidyl-tyrosine, 



with the free amino groups of lysine in the proteins. 

 These normally non -antigenic proteins were thus con- 

 verted into antigens which provoked specific antibodies 

 when injected into rabbits, although rather poorly. 

 This affords further evidence that gelatin and insulin 

 are non-antigenic due to lack of tyrosine and carbo- 

 hydrate and of carbohydrate respectively. Glucosidyl- 

 tyrosine coupled to ordinarily antigenic proteins, such as 

 globulin or albumin, gave very good antigens of a sharp 

 specificity determined by the introduced groups. 

 Harrington also made antigens containing thyroxine by 



