Vol. 44, 1958 BIOCHEMISTRY: BENZER AND F REESE 117 



fold higher, and only half as many induced mutants are mapped; therefore, the 

 occurrence of an induced mutant once represented in the induced set corresponds to 

 roughly 10^ times higher mutability than the occurrence of a spontaneous mutant 

 once represented in the spontaneous set. Hence, for those "hot spots" in the induced 

 set for which no spontaneous mutation has been observed, the mutability with 

 bromouracil is at least 10^ times larger than is the spontaneous mutability. 



Reverse Mutations. — Among the spontaneous rll mutants, not only are forward 

 mutation frequencies different for different sites, but the reversion indexes as well 

 cover an enormous range. Among the 132 spontaneous rll mutants, reversion in- 

 dexes range as high as 5 X lO^^ down to less than IQ-^ For 19 of them, reversion 

 has not been detected. 



The induced mutants, on the other hand, are more homogeneous with respect to 

 reversion index. There is only one mutant of very high reversion index (N 76) and 

 two non-reverting mutants (N 101, N 32). Most of the remainder have reversion 

 indexes of the order of 10 ~l Notably rare are mutants containing larger aberra- 

 tions (i.e., ones that fail to give standard-type recombinants with two or more 

 other mutants that do themselves show recombination). 



Some preliminary experiments have been performed on the induction of reversion 

 of various mutants with 5-bromouracil. For some mutants, in particular, chosen 

 from some of the "5-bromouracil hot spots," increases over the spontaneous rate 

 of reversion by factors as high as 10* have been observed. Some spontaneous 

 mutants show no or only much smaller effects. These experiments are still in 

 progress. 



DISCUSSION 



Clearly, the mutagenic effect of 5-bromouracil is not merely a general enhance- 

 ment of spontaneously occurring mutations. Rather, it is a specific effect. In de- 

 noting the action of a mutagen as specific, one might be comparing the effects on 

 different phenotypic characteristics in the whole organism, different cistrons, or 

 different locations within a single cistron. Were it not for the high resolution of our 

 genetic techniques, it might have been erroneously concluded that 5-bromouracil 

 acts aspecifically, since, among all the induced mutants of the r phenotype, the 

 proportion of rll mutants and even the ratio of A cistron to B cistron mutants are 

 quite comparable to the spontaneous values. It is only at finer resolution that the 

 different effects are revealed, and it is seen that mutations are, for the most part, 

 induced at specific places in the genetic structure. 



Since the proportion of large aberrations and non-reverting mutants is notably 

 smaller among the 5-bromouracil-induced mutations, the induced changes in the 

 genetic structure appear to involve small molecular substitutions rather than large 

 changes of the genome. Further, the more homogeneous properties of the induced 

 mutants with respect to reversion rates indicate that a certain class of molecular 

 transitions may be involved here. 



Under the conditions used, the total frec^uency of occurrence of all rll mutants in 

 5-bromouracil was raised several hundred fold above the spontaneous rate. The 

 increase in rate at specific locations in the genetic structure was much larger. For 

 example, 1 1 occurrences of mutation at one site were observed out of 67 induced rll 

 mutants, while none occurred at that site among twice as many spontaneous rll 



225 



