4 f > 



thousandfold. Under these circumstances, the number 

 of plate-mutants in .1 heavily inoculated plate with 

 limited enrichment should I"- only a negligibli frai tion 

 of the number of mutants inoculated. The number of 

 ultraviolet induced prototroph colonies observed would 

 therefore be expected to be relatively independent of 

 the enrichment. To our surprise, with certain mutants 

 the dependence was found to be extreme. With a 

 washed inoculum of a mutant, almost no ultraviolet- 

 induced prototrophs could be detected unless the 

 medium was enriched with a trace of the factor re- 

 quired by tin parent strain. To look at one instance: 

 an inoculum of 10 8 cells of a tryptophan-requiring 

 mutant yielded 1 to 3 visible back-mutant colonies 

 after three days on minimal agar, and 1 to 6 colonies 

 on agar slightly enriched with 0-01 to 0-5 y ml of 

 tryptophan. The suspension was irradiated until about 

 2 and 50% of the cells survived, and the same volumes 

 were inoculated. On minimal agar, to 2o colonies 

 developed; on slightly enriched agar, 2(>(> to 600! Of 

 these, 25 were pieked at random, ranging from the 

 large to the microscopic. All were prototrophs, growing 

 on minimal agar, though many of the smallest grew 

 very slowly. Qualitatively similar observations have 

 been made with several amino acid, purine, and vita- 

 min requiring mutants. 



It appears that a certain amount of growth by the 

 parent strains is necessary to permit the ultraviolet- 

 induced back-mutants to get started. It is known, 

 however, that small inocula of various bacteria often 

 fail to initiate growth in a medium which is adequate 

 for larger inocula; the small inocula are presumably 

 unable to accumulate enough C0 2 or other essential 

 metabolites. To rule out the possibility that the 

 growth of the parent strain was promoting tin- ap- 

 pearance of back-mutant colonies by some such non- 

 specific mechanism, variously enriched plates were 

 inoculated with an irradiated strain, an unirradiated 

 strain with a different requirement, and a mixture of 

 the two. Prototrophs appeared in huge numbers only 

 from the irradiated strain, and when the medium was 

 supplemented with its growth fat tor rather than that 

 of the companion unirradiated strain. 



We are therefore brought back to the consideration 

 of the phenomic lag, introduced into this work at the 

 outset by the initial failures of the penii illin method. It 

 now appears that a lag occurs not only in exhausting 

 preroutational enzymes, but also in building up new 

 enzymes which the mutated cell is ( apableof construct- 

 ing; to build up the new enzyme the cell requires a 

 complete set of building blocks, including the product 

 of the previously deficient enzyme. In other words, to 

 1. nt the cycle, the pump must first be pruned. This 

 phenomenon stands in contrast to the behavior of 

 certain adaptive fermentative enzymes, which are re- 

 ported to be foi tned by bacteria or yeasts in a nitrogen 

 free medium, in which no net growth can take place. 



I'se 0/ mutants in studying mechanisms <>/ bacterial 

 inhibition 



D-Serine exerts a marked inhibitory effect on the 

 growth of our wild-type strain of /:. coli; many amino 

 acids are able to antagonize the inhibition, but aspartic 

 acid, in contrast, enhances the effect, although by itself 

 it has no influence on growth 1 . Werner Maas, further 

 studying this problem in this laboratory, found that 

 the inhibition is oven ome by extremely low 1 oncentra- 

 tions of pantothenic acid, and by somewhat higher 

 concentrations of /^-alanine. The antagonistic action 

 was apparently non-competitive with pantothenate 

 and competitive with ^-alanine. The results suggested 

 that /^-alanine is the substrate and pantothenate the 

 product of the inhibited reaction. This type of inhibi- 

 tion analysis lias been widely used to determine 

 whether a given antagonist serves .is substrate, pre- 

 cursor, or product of the inhibited reaction. For several 

 reasons, however, which are dis< ussed elsewhere-, it 

 seems difficult to draw rigorous conclusions as to the 

 site of the inhibition, especially in those cases where 

 only a narrow range of concentration of the inhibitor is 

 possible before other reactions become affected. With 

 appropriate mutants, on the other hand, it is possible 

 to dissect out the system under investigation ami 

 produce more direct evidence of the site of inhibition. 

 In addition, mutants show whet her an apparent product 

 of the inhibited reaction, acting non-competitively on 

 wild-type, has done so in "physiological" concentra- 

 tions. 



Mutants blocked at various stages in the synthesis 

 of pantothenic acid were therefore isolated (Fig. 3). A 

 mutant blocked in the synthesis of ^-alanine showed a 

 competitive relation between /^-alanine and D-serine; 

 a mutant blocked in the synthesis of pantoic acid 

 showed inhibition by D-serine when grown on pantoic 

 acid, but none when grown on pantothenic acid; and a 

 mutant blocked in the formation of pantothenic acid 

 grew in proportion to the amount of pantothenic acid 

 present, regardless of the presence or absence of D- 

 serine. It has therefore 1 been demonstrated, more con- 

 clusively than would be possible with wild-type alone, 

 that D-serine (or a product of it) interferes with the 

 conversion of ^-alanine to pantothenic acid'-. Similarly, 

 salicylic acid, which is known to be antagonized bv 

 pantothenic acid 3 , has been found 4 to interfere' with 

 the synthesis rather than the utilization of pantoic 

 acid. Furthermore, several of the amino acids which 

 antagonize D-serine inhibition of wild-type were tested 

 against D-serine inhibition of a mutant unable to 

 synthesize /5-alanine. No antagonism of the inhibition 



102 



