II. REPLICATION OF DNA IN CHROMOSOMES 79 



They clianged a rapidly growing culture of E. cull labeled with N'"' to 

 a medium in which the nitrogen source contained the lighter isotope, N^"*. 

 They took samples at intervals as replication continued. When lysates 

 of these samples of cells were centrifuged in the Spinco Model E and 

 photographed at equilibrium, as shown in Fig. 3, the behavior of the 

 labeled DNA undergoing replication was revealed. A new band of DNA 

 with a density intermediate between the N^^- and N^''-labeled material 

 appeared first. By the end of one cycle of replication all of the DNA 

 was in the band with this intermediate density, while the original mate- 

 I'ial with a high density had disappeared. As growth continued a band 

 corresponding in density to N^'^-DNA began to appear. This N^^-DNA 

 continued to increase in amount, but the band of intermediate density 

 did not disappear. These results indicated very clearly that replication 

 is semi-conservative at the molecular level (Fig. 4). ]\Ieselson and Stahl 

 (1958) made another observation of significance. When the DNA of 

 intermediate density was heated for 30 minutes at 100°C and centrifuged 

 again, it formed two separate bands. Therefore, the DNA of intermediate 

 density was composed of hybrid particles in which the separable strands 

 had different densities. The question raised was difficult to answer. Did 

 the replication involve the unwinding of the two polynucleotide chains 

 of a Watson-Crick double helix? Originally this problem of how two 

 chains with one twist every 34 A could separate was one of major con- 

 cern. Further consideration and some calculations of the energy require- 

 ments and the time that might be required for unwinding have indicated 

 that the problem was more imaginary than real. Levinthal and Crane 

 (1956) showed that the energy requirement and the speed to be 

 expected, if the double helix rotated like a cable, were well within the 

 limits available for cellular processes. More recent experiments and 

 calculations have indicated tiiat no problem exists if the replicating units 

 are similar in size to the isolated particles of DNA (Miller. 1961 ; 

 Geiduschek, 1961). 



Similar results for DNA of mammalian cells were soon provided. 

 Djordjenic and Szybalski (1960) and Simon (1961) used bromouracil as 

 a density label for DNA and obtained evidence for a semi-conservative 

 replication in human cells (HeLa strain) in culture. Evidence was also 



X'^. Each photograph was taken after 20 hours of centrifugation at 44,770 rpm in a 

 .sohition of CsCI; the density of the solution increases to the right. The time of 

 sainphng was measured in units of generation time as indicated on the right, (b) 

 Microdensitometer tracings of the DNA bands adjacent to each photograph. The 

 density of each band can be compared by reference to the two lowest frames which 

 show the .separation of the mixtures indicated on the right. (From Meselson and 

 Stahl, 1958.) 



