In 1958 Matthew Meselson and Franklin Stahl at the California Institute of Technology convinced the scientific community that DNA is reproduced by semiconservative replication. They used density labeling to distinguish between parent strands of DNA and newly copied ones. Historians of science have called this one of the “most beautiful experiments” in biology. For that reason alone, as well as its results, it’s worth taking a close look at it.

275

The key to the experiment was the use of a “heavy” isotope of nitrogen. Heavy nitrogen (¹⁵N) is a rare, nonradioactive isotope that makes molecules containing it denser than chemically identical molecules containing the common isotope ¹⁴N. Two cultures of the bacterium E. coli were grown for many generations, one in a medium containing ¹⁵N and the other in a medium containing ¹⁴N. When DNA extracts from the two cultures were combined and centrifuged in a solution of cesium chloride, which forms a density gradient under centrifugation, two separate bands of DNA formed in the centrifugation tube. The DNA from the ¹⁵N culture was heavier than the DNA from the ¹⁴N culture, so it formed a band at a different position in the density gradient. Figure A in the work with the data portion of Investigating Life: The Meselson–Stahl Experiment shows a photo of the two bands.

Next, Meselson and Stahl grew another E. coli culture in ¹⁵N medium, then transferred the bacteria to normal ¹⁴N medium and allowed them to continue growing. The cells replicated their DNA and divided every 20 minutes. Meselson and Stahl collected some of the bacteria at time intervals and extracted DNA from the samples. You can follow their results for the first two generations in Investigating Life: The Meselson–Stahl Experiment. The results can be explained only by the semiconservative model of DNA replication. The crucial observations demonstrating this model were that all the DNA at the end of the first generation was of intermediate density, while at the end of the second generation there were two discrete bands: one of intermediate and one of light DNA. If the conservative model had been true, there would have been no intermediate density DNA. If the dispersive model were correct, then the DNA would all have been intermediate for the first few generations, with the single intermediate band becoming progressively lighter.

Not long after this experiment, other scientists investigated DNA replication in the presence of cisplatin (described in the opening story of this chapter). Repeating the Meselson–Stahl experiment with cisplatin added to the growth media, they found no change in the density of DNA after several generations—that is, only one DNA band formed in the tubes subjected to centrifugation. From this the researchers deduced that in the presence of cisplatin the two strands of DNA must not be separating during replication. We will discuss more specifically how cisplatin prevents DNA strand separation at the end of this chapter. First, let’s consider the chemistry involved in DNA replication.