46 



tionallv DehYimt B.i< t.-nal Mutants 



of biosynthesis. If the accumulation of normal meta- 

 bolite A in excessive concentration completely blocks 

 a certain enzymic reaction in the production of com- 

 pound B, then the normal concentration of metabolite 

 A might exert a governing effect on that reaction, 

 while a complete absence of metabolite A, resulting 

 from a genetic block, might permit excessive synthesis 

 of compound B, out of proportion to the remainder of 

 the metabolites being synthesized, with resulting ex- 

 cretion of compound B. 



To apply this concept to the present problem, we 

 would postulate a tyrosine precursor which interferes 

 with phenylalanine synthesis, and vice versa. The ex- 

 cretion of phenylalanine would therefore be due to a 

 block early enough in the synthetic chain of tyrosine to 

 cause absence of that tyrosine precursor which normally 

 governs the rate of phenylalanine synthesis, and the 

 same consideration would apply to the phenylalanine 

 mutant which secretes tyrosine. On the other hand, 

 there is one tyrosine mutant which fails to feed phenyl- 

 alanine; its block would occur after the governing 

 intermediate (E in Fig. 4). This scheme is strongly 

 supported by the fact that the excretion of phenyl- 

 alanine by the tyrosine mutant is prevented in the 

 presence of an excessive concentration of tyrosine ; the 

 same is true of the phenylalanine mutant whose ex- 

 cretion of tyrosine is inhibited by an excess of phenyl- 

 alanine or phenylpyruvic acid, which can be sub- 

 stituted for phenylalanine as a growth factor for this 

 mutant. Presumably the excess of the amino acid 

 causes reversal of the normal processes of synthesis, 

 restoring from without the governing compound whose 

 synthesis is genetically blocked. 



This scheme for explaining the output of phenyl- 

 alanine and tyrosine is on speculative grounds, since 

 the intermediates in the synthesis of these amino acids 

 (except for phenylpyruvic acid) are as yet unknown. 

 The concept of normal physiological interaction among 

 separate biosynthetic paths, however, is further 

 strengthened by returning to the better documented 

 isoleucine-plus-valine case. We have isolated a mutant 

 of this type with E. coli whose mechanism appears to 

 be identical with that of the similar Neurospora 

 mutant. In addition, another E. coli mutant has been 

 isolated which responds equally well to isoleucine, its 



a-keto acid, or y.-amino butyric acid. The double- 

 requiring mutant, which cannot use ketoisoleucine, it 

 known to accumulate an inhibitor of valine synthesis; 

 the single-requiring mutant, which can use ketoiso- 

 leucine, and which is fed by the double, must be block- 

 ed earlier ; it would therefore follow, from the reasoning 

 outlined above, that the single-requiring mutant 

 should not form the isoleucine intermediate that gov- 

 erns valine synthesis, and hence might be expected to 

 excrete valine. On testing, this mutant was indeed 

 found to feed a valine mutant heavily, and paper 

 chromatography on the culture filtrate showed a 

 dense spot corresponding to valine. While in this case, 

 as in those of tyrosine and phenylalanine, neither 

 microbiological assay nor paper chromatography alone 

 furnishes complete proof that the compound excreted 

 is the amino acid rather than a related compound, it 

 seems unlikely that the non-specificitv of these two 

 techniques would overlap to give a false answer when 

 used together. 



In several instances, therefore, interference with 

 synthesis of one amino acid leads to excessive synthesis 

 of another. This evidence encourages us to feel that 

 the mutants lend themselves not only to determining 

 the normal steps in biosynthesis, but also to unravelling 

 the mechanisms of integration by which the normal 

 cell determines that its metabolites should be synthes- 

 ized in proper proportions and not wasted. Recent ex- 

 periments have shown that this economy of the 

 organism is upset in diverse ways by a single mutation. 

 A tyrosine mutant, for example, not only excretes 

 phenylalanine, but also, to a smaller degree, feeds the 

 mutants requiring tryptophan, lysine, valine, and 

 leucine. This type of study has just begun, but it is 

 already clear that the interrelations between various 

 biosynthetic paths are complex (Fig. 4). 



Precursors of aromatic compounds 



The mechanism of biosynthesis of aromatic com- 

 pounds has long been a subject of speculation. The 

 availability of mutants with multiple aromatic re- 

 quirements (Fig. 4) offers an opportunity to obtain 

 definite information on this matter. A wide variety of 

 aromatic compounds, with single and multiple sub- 



s» 



*S 



phenylpyruvic acid — > phenylal, 



|— ► tryptophan 



99 



