118 BIOCHEMISTRY: BENZER AND FREESE Proc. N. A. S. 



mutants. Thus the probability of mutation at that point, per repHcation of the 

 phage, was raised by a factor greater than twice eleven times several hundred, or 

 of the order of 10^. 



Conversely, note the site at which there were 19 occurrences among 132 spon- 

 taneous r-II mutants, while among the 67 induced mutants none were observed. 

 Therefore, 5-bromouracil had little, if any, positive mutagenic effect at that point. 

 The disappearance of the spontaneous "hot spot" does not, of course, mean that 

 spontaneous mutations at that point were suppressed but simply that they were not 

 increased in proportion to other mutations. 



Only three sites (indicated by dotted lines in the figure) are represented in both 

 sets of mutants. In each of those cases, there was only a single occurrence in one 

 of the sets, so that the ratio of induced to spontaneous frecjuencies cannot be re- 

 liably computed. The mutation rate seems to have been truly increased by 5- 

 bromouracil in 2 of the cases (2 and 7 occurrences in the induced set) because one 

 occurrence among the induced series represents a thousandfold greater probability 

 of mutation. In the third case, there is, in the induced set, a single occurrence of a 

 highly revertable mutant that appearaed nine times in the spontaneous set. This 

 single event probably belongs to the small background of spontaneous mutations 

 among the induced ones {ca 1 or 2 in 100). 



The methods used in the isolation of the two sets of mutants were quite different, 

 and one cannot be sure whether factors other than the addition of 5-bromouracil 

 could have affected the results. Although sulfanilamide alone does not appreciably 

 increase the total mutation rate, the possibility that the distribution of mutations 

 might change has not been excluded. Sulfanilamide, as an analogue of p-amino 

 benzoic acid, inhibits the formation of folic acid, thereby inhibiting nearly all the 

 methylation and hydroxymethylation steps (Cohen and Earner). i« In our sul- 

 fanilamide medium, nearly all the major chemicals containing the methyl or hy- 

 droxymethyl group are added except the deoxyribonucleotides of thymine and 5- 

 hydroxymethylcytosine, which thus are expected to be deficient. The deficiency 

 in thymine facilitates the incorporation of 5-bromouracil into DNA. Whether 

 the deficiency in 5-hydroxymethylcytosine enhances the probability of the false 

 incorporation of 5-bromouracil into a hydroxymethylcytosine site of the phage DNA 

 remains to be seen. 



One is not in a position, from these experiments alone, to reach a clear conclusion 

 as to the molecular mechanism of mutagenesis; thus, at this stage, it cannot be de- 

 cided whether or not the tautomeric shift of the DNA bases from the keto- to the 

 enol- form is mainly responsible for the production of point mutations, as suggested 

 by Watson and Crick. ^ If it is assumed that only two kinds of nucleotide pairs are 

 present in DNA (i.e., adenine— thymine, and guanine— 5-hydroxymethylcytosine). 

 the existance of hot spots of spontaneous mutation and the appearance of different 

 hot spots with 5-bromouracil would suggest that not every nucleotide pair of a 

 given type mutates with the same probability. Rather, the mutability of a nu- 

 cleotide pair would have to depend upon its position. 



In any case, the striking results of this preliminary investigation indicate that it 

 would be fruitful to pursue this line of investigation, using a range of mutagenic 

 substances in systems where the chemical events are under proper control. 



226 



