THE FOLIC ACID COMPLEX 



synthesised from 2:4:5: triamino - 6 - hydroxy - pyrimidine and 

 glyoxal. 



The aromatic amine fraction on acid hydrolysis yielded _/)-amino- 

 benzoic acid and an a-amino acid, subsequently identified as glutamic 

 acid. Microbiological assay indicated that three moles of the latter 

 were liberated from each mole of amine. 



When the fermentation L. casei factor was treated with sulphurous 

 acid, a pteridine fraction and an aromatic amine were formed. The 

 former contained a carbonyl group and, on standing in dilute alkali 

 solution in the absence of air, underwent a type of Cannizzaro reaction, 

 yielding 2-amino-4-hydroxypteridine-6-carboxylic acid and 2-amino- 

 4-hydroxy-6-methyl-pteridine. The latter was identified by com- 

 parison with a sample prepared by decarboxylation of 2-amino-4- 

 hydroxy-6-pteridine-acetic acid, which in turn was prepared from 

 2:4: 5-triamiino-6-hydroxy-pyrimidine and methyl yy-dimethoxy- 

 acetoacetate. 



Final proof that the methyl group of 2-amino-4-hydroxy-6-methyl- 

 pteridine was in the 6-position was obtained by degradation of the 

 compound to 2-amino-5-methylp5n-azine by the method of J. Weijlard 

 et al.^ 



The liberation of the aromatic amine and the pteridine moiety 

 indicated that the latter was attached to the amino group of ^-amino- 

 benzoic acid. The necessity for oxygen in the alkaline cleavage of 

 the L. casei factor, with other considerations, suggested the presence 

 of a methylene group, and this view was supported by the fact that 

 cleavage of N-benzyl-_^-aminobenzoic acid with alkali was accelerated 

 by the presence of oxygen. 



Proof of the above structure for the liver L. casei factor was 

 obtained by synthesis (see page 474). 



The evidence so far discussed throws no light on the manner in 

 which the three glutamic acid residues are linked together in the 

 fermentation L. casei factor. There are five theoretically possible 

 isomers of the triglutamate and two possible isomers of pteroyldi- 

 glutamate. Mowat et al.'^ synthesised pteroyl-a-glutamylglutamic 

 acid and pteroyl-ay-glutamyldiglutamic acid and showed that they 

 had only a slight activity towards L. helveticus and 5. faecalis R, so 

 that the latter was not identical with the fermentation factor. Pteroyl- 

 y-glutamy]glutamic acid, however, had 60 to 70 % of the activity of 

 pteroylglutamic acid on S. faecalis R and L. helveticus whilst pteroyl- 

 y-glutamyl-y-glutamylglutamic acid, prepared from the triethyl ester 

 of pteroyl-y-glutamylglutamic acid and the acid chloride of a-ethyl 

 carbobenzoxy-glutamate (page 476), had a low activity for S. faecalis 

 R and a high activity for L. helveticus, thus resembling the fermentation 

 factor, with which it is probably identical.^ 



472 



