202 THE BIOCHEMISTRY OF B VITAMINS 



purines (section I) ; in the presence of amino acids the inhibition produced 

 by p-aminobenzoic acid inhibitors can be prevented by thymine and 

 purines (section D) ; for some bacteria, 4(5)-amino-5(4)-imidazolcar- 

 boxamide is partially effective as a purine substitute, and the presence of 

 formic acid enhances its activity 277 ; this amine accumulates when sul- 

 fonamides block the synthesis of purines. 290 Folic acid can cause a hema- 

 topoiesis in certain macrocytic anemias; thymine and nucleotides of 

 purines and pyrimidines also have produced a similar response (p. 414) . 



Mention has been previously made of instances in which the nutritional 

 requirements of certain bacteria for a particular vitamin (biotin, p. 173 

 or pyridoxal, p. 184) can be completely satisfied by supplying only the 

 products of the reactions for which the coenzyme of the vitamin is 

 required, and none of the vitamin itself. This phenomenon was first 

 demonstrated with folic acid. In a medium containing purines and amino 

 acids, thymine can substitute for the folic acid required for growth of a 

 lactobacillus. No synthesis of folic acid could be detected in the cells 

 grown upon this medium. 301 



Composition of the Coenzymes. Of a number of questions still un- 

 answered concerning these vitamins, one of the most perplexing is: Are 

 p-aminobenzoic acid and folic acid used to form the same coenzymes? 

 An independent requirement for both p-aminobenzoic acid and folic acid 

 has never been demonstrated; consequently one cannot, on the basis of 

 nutritional requirements, imply the existence of separate coenzymes for 

 these two vitamins. Nothing is known concerning the chemical nature of 

 their coenzymes; hence, it is impossible to use chemical composition as 

 a criterion for deciding the question. No coenzymatic activity has yet 

 been associated with the various combined forms of folic acid and 

 p-aminobenzoic acid; as a result, a quantitative comparison of the rela- 

 tive distribution of "combined folic acid" and "combined p-aminobenzoic 

 acid" cannot provide a solution based upon the distribution of their 

 actual coenzymes. Nor can a decision be based upon information con- 

 cerning the biological activities of the two vitamins under various con- 

 ditions; the relative responses produced by a vitamin, its coenzymes, 

 and compounds of intermediate complexity depend entirely upon the 

 type of biological system used; and, except in the case of coenzymatic 

 activity in cell-free systems, the responses of the various derivatives 

 have no predictable relationship to their chemical complexity. 



Consequently, a definite statement as to the structure of the coenzymes 

 derived from these two vitamins and their identity or relationship to 

 each other must be delayed until isolated systems can be employed to 

 establish the identity of the specific cofactors. 



