118 MUTATIONS 



to revert to the wild type, and those that cannot revert. Mutants of 

 the first class are demonstrably the result of mutation at a single site 

 of a gene locus, whereas the nonreverting mutants are due to mutation 

 affecting two or more adjacent sites. These so-called multisite mutants, 

 since they show properties characteristic of the deletion mutants 

 found in higher organisms, are assumed to originate by deletion of a 

 small segment of the gene string, and I shall refer to them as deletions. 

 As a rule, the frequency of deletions is low, about 4 per cent, as com- 

 pared with the frequency of single-site mutations. The only exception 

 detected so far is at the cysC (cystine-controlling) region, where close 

 to 40 per cent of the spontaneously occurring mutations are deletions. 

 All the deletions above 4 per cent found in the cysC region form a 

 uniform group. Detailed studies of 44 of them showed that they are 

 very similar to one another with regard to the positions of the two 

 ends, one being in the left third of the cysC region and the other either 

 at the right end of cysC or beyond that end of the locus. The available 

 evidence makes it seem probable that the sectors within which the two 

 ends of these deletions are located are duplications, and that the 

 deletions originated during replication of the gene string by the forma- 

 tion of a tight loop and subsequent exclusion of the portion within the 

 loop from the newly formed string. The other deletions are variable in 

 length and cover different parts of the cysC region. Since the additional 

 deletions are longer than the others, they are called "long deletions," 

 whereas the others are classified as "short" even though some are of 

 considerable length. By dividing the deletions into two classes I do not 

 wish to imply that their modes of origin are different. It seems likely 

 that all deletions originate by the subtraction of nucleotide pairs be- 

 cause of "skips" during replication, due either to sliding between 

 adjacent nucleotides of a pair or to the formation of loops. In the cysC 

 region a duplication of considerable length would be responsible for 

 frequent synapsis of duplicated segments and the formation of loops. 

 Work on this problem is still in progress. The results available through 

 January 1961 are given in Table 2. Additional experiments are now 

 being made with sodium nitrite and UV. The data show clearly that 

 2-aminopurine induces only single-site mutants, whereas all the other 

 mutagens induced long deletions as well. It appears that, of the five 

 mutagens tested, sodium nitrite is the most potent inducer of deletions. 

 Short deletions are so rare that the mutagenic differences shown in the 

 table are not significant. 



Freese: What occurs more frequently with X-rays, long or short 

 deletions? 



