192 S. S. COHEN 



their arrays of chemical structures and interconnecting bonds, codes which 

 orient in some more or less direct mechanism the specific assembly of nucleic 

 acids like themselves and specific proteins, as well. 



A. DNA Duplication and Partition 



In the duplication of DNA, the Watson-Crick hypothesis postulates that 

 the array of nucleotides in one helical chain of a duplex determines the 

 ahgnment of complementary nucleotides, which then combine to form 

 another helical chain to complete the duplex. A number of chemical facts 

 have appeared which present some difiiculties for this duplication mechanism. 

 It is known that in the DNA of animal and plant cells, both 5-methyl cyto- 

 sine and cytosine fill the cytosine complement of the polynucleotide chain. 

 Furthermore, 5-methyl cytosine is usually associated with guanine in di- 

 nucleotide sequences. The problem of forming a chain in which these bases 

 are specifically placed has evoked the hypothesis that in some instances the 

 DNA precursors maybe di- or polynucleotides (Crick, 1957b); such a postulate 

 has been challenged in turn (Chargaff, 1957). 



In T-even phage DNA, cytosine is completely replaced by 5-hydroxy- 

 methyl cytosine (HMC). However, in these viral DNAs, the hydroxymethyl 

 group is glucosylated to varying degrees, depending on the inheritance of the 

 phage. Thus, although bacteriophage T4 appears to contain only a mono- 

 glucosylated HMC, T6 contains HMC nucleotides with 0, 1, or 2 glucose 

 moieties (Cohen and Lichtenstein, unpubHshed data). There is no indication 

 how the existence of a 6-hydroxyl on complementary guanine can select 

 among these tliree types of HMC nucleotides, if indeed such a selection is 

 essential in the determination of genetic specificity of T6 DNA. 



Watson and Crick (1953) have suggested that the synthesis of the comple- 

 mentary chain to form the duplex may occur along both unwinding chains, 

 in a kind of zipper action at the unwinding point. One of the difiiculties in- 

 volved in this hypothesis is the difficulty of separating chains, since these are 

 wound around each other in many turns. A number of ways of separating 

 the chains has been suggested (Delbriick and Stent, 1957), including unwind- 

 ing by a variety of methods or by breaks and reunions. In the latter case, 

 one might even imagine whole lengths of one chain digested away to permit 

 synthesis on the exposed chain; however, this is difiicult to reconcile with 

 the observed conservation of DNA. 



A proposal of Delbriick (1954) conceives of breaks and rejoining to occur 

 in limited segments of the duplex during the replication process. However, 

 his mechanism involves a thorough mixing of sections of parental and 

 daughter nucleotides in the same chain, and recent data appear to exclude 

 this possibility (Meselson and Stahl, 1958). 



