V. MOLECULAR MECHANISM OF MUTATIONS 215 



the genetic map for which recombinants are still detectable experi- 

 mentally or possible molecularly. The unit of recombination may be as 

 small as two adjacent nucleotides, since Helinski and Yanofsky (1961) 

 have found two very closely linked but recombining mutations of the 

 tryptophan synthetase region for which the same amino acid is changed 

 in the corresponding protein. The condition of reversibility is introduced 

 in the above definition since practically all mutants of small genetic 

 extent do revert and since one imagines that the change of a single 

 nucleotide pair can revert occasionally. Yet, the reversion frequency of 

 some mutations (e.g., deletions, insertions) may be unobservably small 

 in the genetic system used. The lack of reversion does not prove that 

 more than one nucleotide pair has been altered. In fact, some rll mutants 

 were thought to be irreversible until their reversion was observed by 

 induction. Conversely, a mutation should not be called a point mutation 

 merely because it reverts. Unless one knows that it has been induced by 

 a chemical which induces nothing but point mutations in other genetically 

 well-studied systems a strong argument cannot be made, for in some 

 cases the revertants may be the result of suppressor mutations (see 

 Section V,B). 



II. Structure, Duplication, and Alteration of 

 THE Hereditary Material 



We wish to determine which molecular groups of the hereditary 

 material are changed in a mutation and how these changes come about. 

 But before we can discuss molecular changes we need to know the 

 molecular structure of the hereditary material itself, i.e., the molecular 

 structure of the chromosome and the mode of its replication. 



A. structure and replication of DNA AND RNA 



The molecular structure of DNA, as proposed by Watson and Crick 

 (1953), is now well established, thanks to the careful studies by Wilkins 

 (1956), Spencer (1959), and Miles (1961). Watson and Crick (1953) 

 also proposed that DNA duplicates by the separation of the two com- 

 plementary strands and the formation of an exact new complement along 

 each parental strand. This hypothesis is supported by studies with 

 density labeling (Meselson and Stahl, 1958), with autoradiography (see 

 Taylor, Chapter II), and with DNA synthesis in cell-free systems (see 

 Bessman, Chapter I). I shall regard this mode of DNA duplication as 

 an established theory. 



Single-stranded, information-carrying DNA and RNA probably repli- 

 cate also by such a complementary base-pairing mechanism. Hence most 



