STRUCTURAL AND CHEMICAL ARCHITECTURE OF HOST CELLS 193 



In the experiments of Meselson and Stahl, E. coli was grown in a medium 

 containing Ni^,and then transferred to an N^* medium to continue its growtli. 

 The culture was sampled (1) when the bacteria were fully labeled in N^^; (2) 

 when they had divided once in N^*; and (3) when growth continued for a 

 number of divisions in N^^. DNA samples were isolated from each of these 

 types of bacteria and analyzed by a sensitive density-gradient ultracentri- 

 fugation technique capable of distinguishing N^^- and N^^-DNA. In the fully 

 labeled bacteria, only N^^-DNA was found. After one division, only one 

 species of DNA was obtained containing equal amounts of N^^ and N^*, 

 impHcating the involvement of both parental chains in the formation of both 

 daughter duplexes. However, in aU subsequent divisions, two species of DNA 

 were obtained, one containing N^* alone, and one contammg equal amounts 

 of N^^ and N^^. Thus, it appeared that single chains remained intact from 

 generation to generation, or, if parts of a strand were broken during duphca- 

 tion, they rejoined to form the original strand. 



Although these experiments tend to support the picture of the duplication 

 of DNA as projected by Watson and Crick, it is not yet entirely certain that 

 such dupHcation does not involve the intermediation of a second template. 

 A direct dupHcation mechanism is also supported by the experiments of 

 Hershey and Melechen (1957) on phage DNA production in the relative 

 absence of protein synthesis. However, it may be suggested that if DNA can 

 orient the synthesis of a specific protein or KNA, it is conceivable that these 

 structures in turn may provide a template for the synthesis of DNA chams. 

 The data to support the latter possibihty are very sparse and stem primarily 

 from an experiment on the effects of infection with P^^ phage containing very 

 highly radioactive P^'- (Stent and Fuerst, 1955). When P^^-labeled cells were 

 infected in P^- medium with P^- phage, on immediate storage at — 197°C. 

 the ability of the infected ceUs to produce phage could be shown to decrease 

 as a function of radioactive decay of the P^^ within the phage. After a few 

 minutes of intracellular phage development, however, the abihty of these 

 ceUs to produce phage became independent of radioactive decay. One 

 hypothesis to explain this result suggests that the code for phage production 

 was transferred to moieties free of decaymg P^^, perhaps to protein. 



A number of workers have attempted to implicate protein in the replica- 

 tion process in another way. The X-ray analyses of DNA, as summarized by 

 Wilkins (1956) and Crick (1957a), have revealed a configuration (the B con- 

 figuration which exists in vivo) in which there are two smooth hehcal grooves, 

 one shallow and the other deep. In nucleoprotamine, the polypeptide chain 

 of protamine is regarded as wmding hehcaUy in the shallow groove around 

 the DNA molecule, which maintains the double hehcal structure, as presented 

 in Fig. 28a. The known heterogeneity of histone necessarily complicates the 

 determination of structure of nucleohistone, but a somewhat similar view of 



