UNION OF ANTIBODY WITH ANTIGEN 123 



strength of the antibody-antigen bond is greatly decreased if the 

 hapten or antigen combining group is sHghtly changed in shape. 

 This is shown by work with haptens of known chemical constitution, 

 such as the experiments discussed in Chapter 1, and by measurements 

 of the bond strength for groups of related haptens, discussed below. 

 The strong influence of shape suggests that close contact between the 

 various atoms of the combining group of the antibody and the atoms 

 of the haptens or antigenic combining group is necessary for a 

 strong antibody-antigen bond. Such closeness of contact accords well 

 with the suggestion, made by Hooker and Boyd (1941), Pauling and 

 Pressman (1945) (Fig. 2-12), and Karush (1956), that the com- 

 bining group of the antibody may in fact be a cavity into which the 

 hapten or antigen combining group fits snugly. Close fit would make 

 the van der Waals forces strong, and any change in the hapten or 

 antigen combining group that lessened that fit would markedly weaken 

 the strength of the bond, which is precisely what we observe. 



Although in the two systems studied by Singer the non-coulomb 

 forces (which, if the argument in the preceding paragraph is valid, 

 may be second in importance) were thought to account for only 

 about half the strength of the antibody-antigen bond, there are 

 cases where the non-coulomb forces presumably account for the 

 entire bond. strength. These cases apply to antigens which do not con- 

 tain positively or negatively charged groups in their specifically re- 

 active portions. Good examples of such antigens are provided by the 

 blood group substances (Chapter 7). Here, no positive or negative 

 groups are present, at least not in the portions responsible for the 

 antigenic specificity. Yet the blood group antigens combine firmly and 

 typically not only with antibody but with the blood group-specific 

 plant proteins I have called lectins (Chapter 6). These reactions have 

 been studied quantitatively (Kabat, 1956; Boyd, Shapleigh, and 

 McMaster, 1955). Karush (1958) believes that the forces between 

 antibody and carbohydrate antigens are mainly hydrogen bonds. 



Wurmser and Filitti-Wurmser (1950) suggest that the combining 

 energy of the isohemagglutinins with their receptors on the human 

 erythrocyte is equivalent to that of about four hydrogen bonds or 

 twenty van der Waals bonds. Before we can discuss such quantitative 

 estimates further it will ])e necessary to go into some of the concepts 

 of thermodvnamics. 



