THE BASIS OF ANTIGENICITY 259 



The Basis of Antigenicity. 



The distinction between haptens and antigens brings us to the second of our 

 inquiries, the basis of antigenicity. It seems likely that the characters determining 

 full antigenic potency of a substance are high molecular weight, and the presence 

 "on the molecular surface of groupings that are distinctive either by reason of their 

 molecular shape, or by reason of high electrochemical activity of certain sub-groups 

 within the group. There are no rigid criteria by which we may judge whether 

 a substance is likely to be antigenic or not. There are only a number of features 

 associated with antigenicity. 



It is sometimes stated that substances are antigenic because they are " foreign " to 

 a given species of animal. This criterion is not very helpful. If we define " foreign " 

 as " harmful " to the animal, we shall cover antigens like snake venom or diphtheria 

 toxin, but not non-toxic antigens like ovalbumin. If we define it as " derived from a 

 dififerent species of organism," we shall expect, for example, insulin from the ox to be 

 antigenic in the rabbit, which is not the case. If, quite unjustifiably, we define it as 

 " not normally present in the tissues of the animal," we assume either that we have a 

 complete specification of the normal constituents of these tissues, or that we can always 

 recognize an abnormahty ; and we shall find in a number of cases that our only criterion 

 of abnormahty of a substance is the response of the tissues to it as an antigen. That 

 is, the term " foreign " can be correctly used to describe all antigens only when it is synony- 

 mous with the term " antigenic." 



We have already noted that antibody response may be conditioned by various 

 distinguishable groupings on the surface of the antigenic particle, and we may 

 reasonably expect that antigenicity itself will be determined by certain combina- 

 tions of chemical groups. These groups, and combinations of them, are called 

 antigenic " determinants," but in seeking to relate antigenicity to the determinants 

 of an antigen we must not be misled by a too restricted notion of a determinant. 

 To call atoxyl a determinant in atoxyl-azo-protein does not necessarily imply an 

 additive effect when atoxyl is linked to the protein. The true determinant is 

 atoxyl in certain positions on the surface of the protein. The atoxyl covers some 

 native groups, or renders them in some way inaccessible to antibody ; and the 

 reactivity of other surface groups may be altered by the presence of the new group. 

 Moreover, the determinant effect that we attribute to atoxyl may depend on its 

 electrochemical and spatial relations with the rest of the surface, whether antigenic 

 or not. 



We may illustrate this distinction by reference to the determinant effect of 

 organic arsenic compounds. Singer (1942) prepared antisera to arsenic-ox globulin, 

 and used it to protect mice against lethal doses of the synthetic antigen. Suc- 

 cessive adsorption of the sera with arsenic-rabbit globulin and ox-globulin did 

 not remove all protective power ; clearly, antibodies to the protein-determinant 

 complex were left in the serum. 



The sphere of influence of the determinant will clearly depend on the chemical 

 structure of the antigen. A protein antigen like horse globulin could not be spread 

 in a unimolecular film without losing reactivity with antibody (Danielli, Danielli 

 and Marrack 1938). Here the relationship of amino-acids in the complete protein 

 molecule was a determinant of specificity. On the other hand, the films of nucleo- 

 protein and other fractions of streptococci spread in thin expanded films on the 

 surface of slides still retained a conspicuous degree of specific reactivity (Bateman, 

 Calkins and Chambers 1941, Chambers, Bateman and Calkins 1941). 



