H. GIDEON WELLS 707 



in the positive experiments such separation has not been demonstrated. When the 

 lipoids are obtained in a high state of purity they have been found to be non-anti- 

 genic. Therefore, it is difhcult to dismiss the opinion that the observed antigenicity 

 of lipoid preparations depends upon admixed proteins, especially when we consider 

 the infinitesimal amounts of protein that may serve as antigens, as pointed out in 

 previous paragraphs. Lipoids readily take up proteins, and there are experiments 

 suggesting that possibly the lipoidal admixture increases the antigenic efficiency of 

 the proteins. 



At present we can say that the antigenic capacity of pure lipoids has not been 

 established. It may be that plants, bacteria, or animal species far removed from the 

 mammals may have lipoids suf^ciently foreign to the mammals to incite antibody 

 formation in them, but such an occurrence has not been conclusively demonstrated.' 



Possibly complex carbohydrates in colloidal state may also be foreign enough to 

 incite antibody formation, for, as Pryde has emphasized, it is possible to secure an 

 enormous variety of carbohydrates. Even as the score of known amino acids of pro- 

 teins may be built up into an incredibly great number of combinations, sugar mole- 

 cules may be linked together in a vast number of ways. Two molecules of J-glucose 

 may combine in at least twenty-eight ways, and Pryde calculates that 12 mols of 

 glucose could form somewhere in the order of 10'^ possible different carbohydrates. 

 However, as yet there has been adduced little evidence that carbohydrates do serve 

 as antigens. Even the bacterial carbohydrates, with their strikingly specific reaction 

 with immune sera, have not been found to be antigenic. Ford has reported that by 

 immunizing with extracts of the poisonous fungus, Amanita phalloides, antisera are 

 obtained which neutralize the hemolytic poison of Amanita, and this hemolysin seems 

 to be a glucoside. As yet no one has repeated this experiment, which stands as the 

 chief example of a possible non-protein antigen. 



ANTI-LIPOID SERA 



Further support to the idea that antigenic lipoidal mixtures owe their antigenic 

 activity to associated proteins is furnished by demonstration of the readiness with 

 which mixtures of non-antigenic lipoids with proteins produce immune bodies react- 

 ing with lipoids.^ Landsteiner and Simms^ found that the isolated lipoidal element of 

 so-called "heterogenetic antigens" is itself non-antigenetic, but when added to normal 

 foreign serum the mixture functions as an efficient heterogenetic antigen. If the 

 lipoid is injected into a vein of one ear of a rabbit, and the serum into the other ear 

 at the same time, little or no effect is obtained, proving that the lipoid must be in 

 some sort of combination with the antigenic protein to have any effect on the anti- 

 body production. Rabbit serum mixed with lipoid will function to some extent as a 

 heterogenetic antigen when injected even into rabbits, although mixtures of foreign 

 serum and foreign lipoids are more potent. The active lipoids in these experiments 



'See Schmidt, H.: Zlsc/ir.f. InimiiHitdlsJorsch. u. cxper. Tl/ernp., 38, 511. 1Q24; Klopstock, A.: 

 Klin. Wclinschr., 6, 119. 1927. 



-Review by Klopstock, A.: Ztschr.f. Immtinitdtsjorsch. 11. ex per. T/ierap., 48, 97. 1926. 



3 Landsteiner, K., and Simms, S.: /. Exper. Med., 38, 127. 1923; see resume in Klin. Wchnschr., 

 6, 103. 1927. 



