7i6 THE CHEMISTRY OF ANTIGENS 



be possible that such a minute amount of protein can be exclusively responsible for 

 antibody production. If so, much of the reported work with ''non-protein antigens" 

 will need reinterpretation, since these amounts are too small to be detected by the 

 usual methods, in the concentrations used. That such amounts of protein can act as 

 antigens may be questioned by many. This point can be settled only by careful 

 quantitative work with pure materials. 



STUDIES WITH PURIFIED MATERIALS 



Extensive investigations on the antigenic properties of purified proteins have 

 grown out of the generally accepted idea that only proteins can act as true antigens. 

 Most of this work has been done with easily available proteins such as casein,^ egg 

 albumin,^ the serum proteins,^ and the plant proteins such as zein, gliadin, hordein, 

 edestin, glutenin, and legumin.^ A comparison of their physical properties and 

 chemical composition with their antigenic activity as shown by the precipitin, com- 

 plement fixation, or the even more delicate anaphylaxis reactions has revealed some 

 interesting facts. A substance may be chemically a protein and yet not be antigenic; 

 as, for example, gelatin which is, however, not a natural but a derived protein, and 

 also proteins that have been racemized.^ But nearly all whole native proteins are 

 antigenic, though by no means equally so. They must be soluble in the tissues, and 

 they must be foreign to the animal into which they have been introduced. A study 

 of the chemical composition of many proteins has shown that those that contain 

 aromatic amino acids are antigenic, even though they may lack others, and that the 

 presence or absence of the diamino acids apparently has no influence upon antigenic 

 activity. These findings have led to the suggestion that the aromatic radicals of the 

 protein molecule may be important in determining its antigenic activity, and that the 

 diamino acids are not of importance. Most of the evidence at hand indicates that 

 combinations of amino acids less than the whole protein molecule are not antigenic. 



Whether or not lipoidal substances can act as true antigens has been the subject 

 of much discussion and experimentation. Many positive results have been reported 

 with such materials. Much of this work has been done with alcoholic or ethereal 

 extracts, and the possibility of small amounts of protein being present cannot be 

 excluded. Apparently some protein-lipoid combinations are more powerful antigens 

 than the protein alone, and produce antisera reacting with both protein and lipoid.'' 



There is much more variation in the structure and distribution of carbohydrates 

 than in fats. The number of possible combinations of the atomic groups which may 

 enter into their composition is comparable to that of the proteins, and theoretically 



• Wells, H. G., and Osborne, T. B.: op. ciL, 29, 200. 192 1. 

 = Wells, H. G.: /. Infect. Dis., 5, 449- iQoS. 



3 Hunter, A.: loc. cit.; Dale, H. H., and Hartley, P.: loc. cit.; Doerr, R., and Berger, W.: 

 loc. cil. 



•t Wells, H. G., and Osborne, T. B.: op. cit., 8,66. 1911; 12, 341. 1913; 17, 259. 1915; 19, 183. 

 1916; Wells, H. G.: ihid., 5, 449. 1908. 



s Ten Broeck, C: J. Biol. Chem., 17, 369. 1914. 



* The relation of the protein and the lipoid to the antigenic activity of such mixtures is more 

 fully discussed in chap. 1 of this volume. 



