V. MOLECULAR MECHANISM OF MUTATIONS 255 



Such a situation may hold, e.g., for a functionally important amino acid 

 in the "active center" of the enzyme or the presence or absence of a 

 proline which produces a kink in the protein. In these cases only a back 

 mutation would return full functional activity. On the other hand, many 

 mutations may cause a change of the enzyme which can be counteracted 

 also by a second mutation at some other genetic site. Such a suppressor 

 mutation may return the functional property of the enzyme partially or 

 even completely. If a given mutagen should induce this second base pair 

 change more frequently than the back mutation, most induced revertants 

 would contain both the original mutation and an intragenic suppressor 

 mutation. 



The question now is, how can one distinguish intragenic suppressor 

 mutations from back mutations? If the functional property is restored 

 only to a limited extent this distinction is easy because one observes a 

 different phenotype. For example, many rll mutants plated on bacteria 

 K produce both standard type and tiny plaques ; at least the latter must 

 be due to suppressor mutations. Profiavin mutants represent an extreme 

 case since most revertants show plaques different from the standard 

 phenotype. Practically all of these revertants are apparently caused by 

 suppressor mutations and back mutations are very rare. Brenner et al. 

 (1961) have proposed that profiavin induces only deletions or insertions 

 of one base pair which rarely back mutate. Crick et al. (1961) have 

 employed these mutations in order to provide some evidence for their 

 coding theory. 



The revertants can be examined further for their functional proper- 

 ties by plating them on different bacterial hosts of type K and B; thus 

 one can distinguish partial revertants from revertants with standard 

 phenotype. Occasionally one observes revertant plaques so similar to the 

 standard type that a clear distinction is difficult; in most cases such 

 phages have not been used for further analysis. 



The next step is to determine whether any of the remaining revertants 

 of standard phenotype still contain both the original rll mutation and a 

 suppressor mutation. To this end one backcrosses revertant to standard 

 type phages and measures the frequency of rll mutants before and after 

 the cross. Many such crosses have been made, leaving out all unclear 

 cases of intermediate plaque types, and no evidence for further sup- 

 pressor mutations was found in these cases (Freese, 1959b; Freese et al., 

 1961b; E. B. Freese, 1962). Thus it would seem that most revertants of 

 genuine standard phenotype arose by back mutations. However, the 

 significance of the backcrosses is limited by the fact that each standard 

 type and revertant stock already contains spontaneous r mutations ; the 

 crosses thus permit one to state only that the sujijiressor mutation cannot 



