48 



B. D. Davis: Studies on Nutritionally Deficient Bacterial Mutants Isolated by Means of Penicillin [Experientia Vol. VI/2] 



restored to optimal by the addition of PABA; others 

 did not. Similar studies were carried out with mutants 

 requiring 6 amino acids, purines, biotin, and panto- 

 thenic acid. In all cases but one amino acid, a variety 

 of degrees of back-mutation were observed; in only 

 a fraction of the strains were slow growth rates restored 

 to optimal by addition of the parent's growth require- 

 ment. 



In the absence of the required genetic techniques it 

 has not been possible to demonstrate whether these 

 various degrees of restoration of a nutritional deficiency 

 represent quantitative alleles of the same gene, or 

 mutations in other genes which modify the deficiency. 

 In the cases where the back-mutant is slow either with 

 or without the growth factor, an allelic change seems 

 much less likely than does a mutation of another gene 

 which imposes a rate-limiting alteration of metabolism 

 at the same time that it restores the deficiency. In any 

 event, it is clear that back-mutants, which have lent 

 themselves to quantitative genetic studies, represent a 

 genetically very heterogeneous class. Indeed, it is not 

 evident that any of the back-mutants are truly 

 identical with wild-type. Perhaps they should be 

 understood as "backward" mutants rather than true 

 back-mutants or reversions, having mutated in the 

 direction, but not necessarily to the precise position, 

 of wild-type. One study of back-mutants of Neuro- 

 spora 1 has failed to reveal such frequent and varied 

 partial restorations of a deficiency. Incidentally, in this 

 connection, it should be noted that mutations from 

 wild-type to partial deficiencies of various factors are 

 frequent, with Neurospora as well as with bacteria ; 

 they are not often mentioned in the literature since 

 they have not seemed to lend themselves to genetic or 

 biochemical analysis as well as mutants with absolute 

 requirements. 



Reversion of multiple requirements 



Another genetic question, more essential for bio- 

 chemical investigation, is whether certain multiple 

 requirements arise from a single mutation and hence 

 presumably from a change in a single enzyme. While 

 we have not tried to solve this problem directly by 

 recombination techniques, it has been possible to 

 circumvent it by use of back-mutations 2 . In several 

 instances it has been easy to obtain spontaneous 

 prototrophs, with no growth requirements, from 

 mutants with multiple requirements (e. g. quadruple 

 and double aromatic), and has not been possible to 

 obtain split reversions, with loss of only part of the 

 requirements. From these results it is inferred that a 

 single mutation has occurred and then reverted. In 

 another instance, requiring histidine and the three 



1 N. H. Giles, Jr., and E. Z. Lederberg, Ainer. J. Bot. 35, 150 

 (1948). 



8 R. R. Roepke, quoted by E. L. Tatum, Cold Spring Harbor 

 Symp. Quant. Biol. 11, 278 (1946). 



aromatic amino acids, it has been possible, in appro- 

 priate media (limited histidine, excess of the other re- 

 quirements) to lose the histidine requirement alone, 

 but it has not been possible to lose all four require- 

 ments in one step. It is therefore concluded that this 

 mutant had been altered in two separate steps. Non- 

 allelic reversions from mutant to apparent wild-type 

 in higher organisms have been described in which a 

 mutation of a second gene at a different locus (modifier 

 gene) restores the normal condition. In the absence of 

 a test for allelism it cannot be decided whether our 

 reverse mutations involve a change of the same gene 

 or of another one. It is conceivable that a mutation of 

 a modifier gene might suppress two independent 

 growth requirements. But whatever the genetic 

 mechanism of the reversions may be, one would expect 

 with a double mutant to recover the single back- 

 mutants as well as the prototrophs. Complete failure 

 to isolate the single reversions, together with isolation 

 of the total reversions, is therefore considered excellent 

 evidence for a single mutation producing a multiple 

 requirement. On the other hand, it should be pointed 

 out that the recovery of split reversions is not conclusive 

 evidence for the presence of independent mutations, 

 unless it is accompanied by failure to isolate any total 

 reversions. With the isoleucine-plus-valine mutant, for 

 example, one back-mutant was obtained with a relative 

 requirement for valine (slow growth without it), but 

 no requirement for isoleucine. Since many completely 

 prototrophic back-mutants were also obtained from 

 this strain, a probable mechanism for the split reversion 

 would be incomplete restoration of the isoleucine 

 enzyme, leaving enough residual accumulation of the 

 intermediate to inhibit valine synthesis. 



This stress on the importance of negative as well as 

 positive results is warranted only because of the great 

 efficiency, essentially 100%, with which reversions to 

 nutritional independence can be isolated from huge 

 bacterial populations. 



Delayed phenotypic expression of back-mutation 



The number of spontaneous back-mutant (proto- 

 troph) colonies observed on a minimal medium plate is 

 proportional to the size of the inoculated population. 

 In a medium with limited enrichment, in contrast, the 

 inoculum grows until the population reaches a size 

 which is limited by the amount of enrichment; the 

 mutants observed are therefore largely "plate- 

 mutants" (i.e., those arising during generations 

 occurring on the plate), and their number is, as a first 

 approximation, a function of the enrichment rather 

 than the inoculum size. Following ultraviolet irradia- 

 tion, however, an entirely different situation was en- 

 countered. 



Ultraviolet irradiation increases the frequency of 

 back-mutants among the survivors by as much as many 



IOI 



