Matthew Meselson and Franklin Stahl show how new DNA is made by copying the old.

I'm Matthew Meselson. I'm Frank Stahl. We showed that new DNA is made by copying from the old. New DNA must be made whenever a cell divides — each daughter cell must receive a faithful copy of the parent cell's DNA. Watson and Crick suggested that during replication the DNA strands separate, and each strand acts as a template for synthesizing a new complementary strand. This is called semi-conservative replication, because each of the daughter molecules consists of one "old" strand — from the parent molecule — and one newly synthesized strand. In 1958, we published results that supported this model. As you will see, our experiment made clever use of nitrogen isotopes and "density gradient" centrifugation. We started by first growing E. coli in culture with 15 N, which is a heavy isotope of the nitrogen atom. As the bacteria grew and divided, the 15 N was incorporated into each of the nitrogen bases of DNA — A, C, G, T. We saved a sample from the 15 N-labeled bacteria for later analysis. The rest of the culture was transferred to fresh culture media containing normal nitrogen — 14N — which is lighter than the 15 N isotope. Thus, only 14 N was available for subsequent construction of DNA as the bacteria replicated. Then, we withdrew culture samples from the 14 N each time the bacterial population doubled. For comparison purposes, we prepared a third culture of bacteria that was grown only in 14 N. Next we examined the nitrogen composition of the bacterial DNA from each culture. To do this, we isolated DNA from the bacterial cultures and dissolved the DNA in a strong salt solution. Then we centrifuged the DNA solution at very high speed. Centrifugal force pulls more salt molecules toward the bottom of the tube, leaving fewer and fewer molecules toward the top. The greater the number of salt molecules at any point in the tube, the greater the density of the solution. We expected that DNA molecules composed of heavy 15 N would "sink" further in the salt gradient than a comparable DNA molecule composed of 14 N. This is exactly what we saw in tubes drawn from the 15 N and 14 N cultures. The critical tubes were those containing samples from the culture first grown in 15 N, then switched to 14 N. In sample# 2, taken after the bacterial population (and the DNA molecules) had doubled, the DNA band we saw was exactly midway between the 15 N and 14 N bands. In sample# 3, taken at the next doubling time, there were two DNA bands. One lined up exactly with the 14 N band, while the other was at the midway point. These data perfectly matched the results predicted by the semi-conservative DNA replication model. The heaviest DNA molecules were found in sample# 1, where both strands of the molecule incorporated heavy 15 N. The lightest DNA molecules were found in sample# 4. Remember, the DNA in sample# 4 contain only 14 N because the bacteria were grown in 14 N. In sample# 2, the DNA molecules collect as a band of intermediate density. When the bacterial culture is switched from 15 N to 14 N, any new DNA made must use 14 N. The two heavy 15 N strands of the parental DNA separate, and act as templates for the incorporation of 14 N into new DNA strands. Thus, all the DNA are hybrids with one heavy 15 N and one light 14 N strand. Sample# 3 underwent an additional round of replication, and there are two separate bands in the centrifuge tube. In this case, each hybrid molecule separates into one heavy and one light strand. The heavy parent strand acts as a template for making a complementary light strand, resulting in another hybrid molecule. The light parent strand is also a template for incorporating light nucleotides, but this time resulting in a molecule with two light strands. The intermediate-density DNA and the light-density DNA show up as two separate bands in the centrifuge tube. We didn’t know the details of how DNA replicates when we published our results. Still, it was clear that Watson and Crick’s predictions about the semi-conservative replication of DNA were correct. I’m Arthur Kornberg. While Meselson and Stahl were doing their experiments, I was isolating the enzyme that synthesizes DNA, called DNA polymerase. I made a cell-free system that replicates DNA. I extracted DNA polymerase from E. coli, then added it to a salt solution containing template DNA molecules and the four deoxynucleotides. I included radioactively-labeled thymine as a "tag" to monitor any new DNA molecules that were made. After incubating these components at body temperature, the radioactive thymine turned up in long polynucleotides. This meant that DNA polymerase had incorporated the radioactive thymine, along with the other nucleotides, as it built a new DNA strand using the available template. I also found that replication only occurs when all four nucleotides are present. Omit one, and polynucleotide chains are not synthesized. DNA polymerase also requires intact DNA to serve as a template. Add DNase, which digests the template DNA into pieces, and polynucleotide production is halted. So, as Watson and Crick had predicted, DNA is used as a template to replicate itself. DNA polymerase is the enzyme that makes it happen. Later, we found that cells have not just one, but three different DNA polymerases! It turned out that the one I discovered, DNA polymerase I — is used mostly for DNA repair, and not for DNA replication. DNA polymerase III, isolated by my son Tom, is the major enzyme that replicates DNA in E. coli.

density gradient centrifugation, matthew meselson and franklin stahl, nitrogen atom, isotope