smith: directed mutation 421 



which 15 also responded to the base analogues. Mutants that were 

 affected by one base analogue were also, with one exception, affected 

 by the other. 



Zamenhof (73) has reported that heating (135° to 155° C) of vege- 

 tative cells (E. colij or spores (B. subtilis) in the dried state in vacuum 

 produced mutations at frequencies of 1 to 10 per cent, certain mutants 

 appearing more frequently than others. It was suggested that this 

 difference may be attributed to a differential susceptibility to altera- 

 tion by heat of different loci. 



The first evidence that the base analogue 5-bromouracil induces 

 mutations more frequently at specific sites, distinctly different from 

 those most mutable spontaneously ("hot spots"), 'was afforded by the 

 experimental results of Benzer and Freese (8) with rll alleles of bac- 

 teriophage T4. These experiments were carried out consequent to 

 evidence that 5-bromouracil is incorporated in place of thymine in 

 the DNA of phage and bacteria (16, 74) and that this incorporation 

 causes mutation (42). Subsequently, Brenner, Benzer, and Barnett 

 (9) demonstrated the induction by proflavine of another series of 

 mutations which occur preponderantly at sites of the ill locus dis- 

 tinctly different from those characteristic of the spontaneous and 

 5-bromouracil spectra. The mutant alleles can be mapped by tech- 

 niques of high resolution which permit the positioning of mutant 

 sites in regions of the order of magnitude of a few nucleotide pairs 

 in a DNA molecule. 



Freese (23, 24, 25) has extended these investigations with rll 

 alleles to include mutagenic patterns of additional base analogues to 

 test the reversion frequencies of induced and spontaneous mutations, 

 and particularly to provide a theoretical basis for better understand- 

 ing of the mutation process and mutagenic specificity. Of the addi- 

 tional analogues investigated, 2-aminopurine, 2,6-diaminopurine, 

 and 5-bromodeoxyuridine were found to be mutagenic. The muta- 

 bility spectrum of 5-bromodeoxyuridine 'was essentially the same as 

 that of 5-bromouracil but different from that of 2-aminopurine. Each 

 differed from the spectrum of spontaneous mutants. A striking differ- 

 ence was shown in the pattern of reversion of ill alleles that had been 

 induced originally by different mutagens. Of the mutants induced by 

 base analogues, 95 to 98 per cent were reversible by base analogues; 

 whereas only 14 per cent of spontaneous mutants and 2 per cent of the 



