46 INTRODUCTION TO IMMUNOCHEMICAL SPECIFICITY 



antibody which might have been directed toward the whole hapten 

 GIL. At the same time, evidence was obtained that the anti-G and 

 anti-L of the antiserum produced by injecting the GIL antigen were 

 not quite identical with those produced by injecting G- and L-coupled 

 antigens. 



Carbohydrate Determinants. Experiments with carbohydrate 

 haptens have given similar results. Goebel and co-workers (Avery, 

 Goebel, and Babers, 1932; Goebel, Avery, and Babers, 1934) in- 

 jected conjugated antigens containing monosaccharides and disac- 

 charides as haptens (Fig. Z-7). Coupling was done by way of an 

 aminophenyl group in each case. It will be seen that the cross-reac- 

 tions of the antisera to the disaccharide haptens occurred mainly with 

 the terminal sugars. Inhibition experiments showed that the /'-amino- 

 phenyl glycosides of the terminal sugars (monosaccharides) were 

 nearly as good inhibitors of the anti-disaccharide sera as the /'-amino- 

 phenyl glycosides of the disaccharides themselves were. This again 

 points to a predominant influence of the terminal group of a com- 

 posite hapten on the antibody produced when an antigen containing 

 it is injected. Nevertheless, the fact that the /'-aminophenyl glycosides 

 of the disaccharides were still somewhat better than the disaccharides 

 alone as inhibitors of their corresponding antisera suggested that the 

 anti-disaccharide antibodies were to some extent directed toward 

 the whole hapten and that a carbohydrate hapten could be larger 

 than a disaccharide. 



Kabat (1957) was able to obtain further information on this point 

 by studying the antibodies produced in human beings by injections 

 of dextran, a large-molecule polysaccharide produced by certain 

 bacteria. Dextran appears to be made up entirely of glucose, pre- 

 dominantly connected by 1-6 linkages (Fig. 3-8). With such a 

 simple antigen the possible antigenic determinants are merely one 

 or more glucose units. Finding out how big an antigenic carbohydrate 

 determinant may have a specifically corresponding antibody determi- 

 nant is simply a matter of finding out how many glucose units an 

 oligosaccharide must contain to fill the combining site on the anti- 

 body. Kabat studied this question by measuring the relative inhibiting 

 power of glucose, isomaltose (two glucose units), isomaltotriose 

 (three glucose units), and larger polysaccharides for an anti-dextran 

 serum acting on dextran. The results are shown in Fig. 3-9. It is 



