GENETICS IN MODERN BIOLOGY — BEADLE 405 



production at a molecular level. Can the hypothesis be tested? The 

 answer is yes. In fact, several kinds of experiments can be and have 

 been made to see if the hypothesis agrees with observed facts. 



In one kind of experiment DNA units are labeled with radioactive 

 phosphorus. Each nucleotide has one phosphorus atom, and a certain 

 number of its phosphorus atoms can be made radioactive by growing 

 an organism, say a bacterium, in a medium containing radioactive 

 phosphorus for several generations until it bcomes equilibrated. 

 Then both chains of its DNA molecules will be labeled. If the bac- 

 teria are then allowed to multiply in a medium in which there is no 

 radioactivity, the two chains of each DNA molecule, both labeled, 

 should come apart, each then directing the synthesis of an unlabled 

 partner. The new double molecules should then be labeled in one 

 chain but not in the other. In the next generation the labeled chain 

 should separate from the nonlabeled one. With synthesis of non- 

 labeled partners by these, there should be produced labeled and non- 

 labeled double molecules in equal numbers. The observed results are 

 consistent with this expectation. 



Another way of doing essentially the same experiment is to replace 

 the normal nitrogen atoms of DNA with "heavy" nitrogen, the stable 

 isotope N^^ instead of the usual N" counterpart. DNA molecules so 

 labeled become heavier but not larger. Hence they are denser. DNA 

 containing only W^ can be cleanly separated from that containing N" 

 in an analytical centrifuge cell in which an appropriate density gradi- 

 ent is established. In such experiments it is found that bacteria con- 

 taining DNA fully labeled with N^^, if allowed to multiply once 

 (double in number) in a medium containing only N^*, give rise to de- 

 scendants in which all the DNA molecules are "hybrid" as though one 

 nucleotide chain of the double molecules contained N^'^ and the other 

 N^*. This, of course, is what is predicted by the hypothesis. In a 

 subsequent generation, also in N^* medium, half the DNA molecules 

 are hybrid and half are fully light. Again this is what would be 

 expected if the hypothesis is correct. 



Wliile experiments of this kind do not prove that the Watson-Crick 

 hypothesis of DNA replication is correct, they do strongly suggest it. 



An even more dramatic w^ay of testing the hypothesis is the one used 

 by Professor Arthur Kornberg and his associates, now at Stanford 

 University. They have devised a test-tube system in which there are 

 present the four nucleotides A, T, C, and G as triphosphates, a buffer 

 solution, magnesium ions, and a polymerizing enzyme. DNA mole- 

 cules added to this system appear to be replicated. Is the new DNA 

 like the primer molecules added? One important observation sug- 

 gests it is. The ratio of A:T nucleotide pairs to C: G pairs of the 

 product is like that of the primer DNA. It is not easy to see how this 

 could be if the primer were not being copied in a precise way. On 



