In the 1950s, details of the genetic material, DNA, began to pour out of laboratories. First, James Watson and Francis Crick built a model for the double helical structure of DNA, showing that one strand consists of a sequence of bases that are complementary to the bases in the opposite strand. These scientists also suggested a model for the replication of DNA that would allow a cell to copy its genetic material and pass down exact replicas to daughter cells. Each old strand of the double helix would serve as a template to make a new strand. Although the simplicity of their replication model was compelling, no data yet existed to prove that it was correct.
A few years after Watson and Crick published their DNA structural model, the scientists Matthew Meselson and Franklin Stahl designed an elegant experiment to determine how DNA replicates.
In 1953, James Watson and Francis Crick built their model of the structure of DNA, the principal features of which are still accepted today. DNA consists of two strands that are complementary in base sequences to each other. A cytosine base in one strand always pairs with a guanine base in the other. An adenine base in one strand always pairs with a thymine base in the other.
The complementary nature of the two strands suggested to the scientists a model for DNA replication. They proposed that the old strands serve as templates to make new, complementary strands. The two resulting double helices would each contain one new and one old strand.
There are other models to consider, as well. For example, the old DNA molecule could be preserved and an entirely new DNA molecule could be produced from it. In yet another model, the result of replication would be two molecules with old and new DNA interspersed along each strand.
Although Watson and Crick proposed the model of semiconservative replication, at the time no evidence existed to prove that this model was correct. To solve this problem, the scientists Matthew Meselson and Franklin Stahl designed an experiment to test Watson and Crick's model of replication.
The key to the Meselson Stahl experiment was devising a strategy to distinguish between old versus newly synthesized DNA. They distinguished the two by labeling them with isotopes. They grew Escherichia coli bacteria in the presence of either a "heavy" isotope of nitrogen (15N) or the ordinary "light" isotope, 14N.
After many generations, the DNA in the bacteria contained either the heavy or the light form of nitrogen, but not both. In this example, the nitrogen atoms in a thymine base are labeled with either the heavy or the light forms of nitrogen.
The scientists took samples of each bacterial culture. They processed the bacteria to release the DNA into solution. Equal volumes of the DNA solutions were mixed together, and then this solution was mixed with a concentrated solution of the salt cesium chloride (CsCl). The density of the CsCl was 1.71 grams per cubic centimeter, the same density as DNA.
The tube was placed in an ultracentrifuge capable of high-speed centrifugation (140,000 X g for 20 hours). At this high speed, cesium ions have a tendency to sediment toward the bottom of the tube, forming a higher density solution at the bottom compared to the top. Other substances, such as the DNA, move to the position within the tube that matches their own density. The DNA with the heavy isotope of nitrogen migrates farther down the tube than the light DNA.
With their technique of separating "heavy" versus "light" DNA established, the scientists tested the hypothesis of semiconservative replication. They first grew E. coli for 14 generations in a medium with 15N in the form of NH4Cl as the sole nitrogen source. Growing the bacteria for many generations ensured that all the DNA would be labeled with heavy nitrogen.
At this time, they isolated their first bacterial sample, prepared the DNA, and added the CsCl for centrifugation. At the same time, some of the bacteria were transferred to 14N "light" medium and allowed to continue to grow. From this point onward, newly replicated DNA will be made with the "light" form of nitrogen.
They called their first sample generation zero. After the transfer to "light" medium, a sample was taken every 20 minutes, which is the generation time for E. coli cells growing at their optimal temperature. The DNA from the samples was prepared for high-speed centrifigation.
Meselson and Stahl found the following results. In generation zero, all of the DNA was in the "heavy" form. After one generation, the DNA was neither "heavy" nor "light", but an intermediate density. After two generations, half of the DNA was light and half was intermediate. After additional generations, more of the DNA was in the light form and less in the intermediate form.
Generations 0 through 2 provide enough information to determine which model of DNA replication is correct.
Consider the possibility of conservative replication. According to the model, after one generation, half of the DNA would be light and half would be heavy. However, the experimental results show that all of the DNA is actually in the intermediate form, so we can reject this model of replication.
Consider the possibility of dispersive replication. According to the model, after one generation, all of the DNA would be in the intermediate form, which is indeed consistent with the experimental results. In the model, after two generations, all of the DNA would be in the lighter form. However, the experimental results show that half of the DNA is actually in the intermediate form, and so we can reject this model of replication.
Finally, consider the possibility of semiconservative replication. According to the model, after one generation, all of the DNA would be in the intermediate form, which is consistent with the experimental results. In the model, after two generations, half of the DNA would be in the intermediate form and half would be in the lighter form, also consistent with the experimental results.
The data provide support for the semiconservative model of replication.
The Meselson-Stahl experiment showed that DNA replicates by a semiconservative mechanism. The double helix separates so that each old strand serves as a template for a new strand. Two new double helices result, each containing one new strand and one old strand.
Although Meselson and Stahl performed their experiment a half-century ago, the experiment has a modern quality. Their experiment represented a new way of doing science that is, making predictions based on different models and then determining which prediction most closely aligns with the experimental data. The materials they used are also still in wide use today. Isotopes are used for tracking atoms and molecules, CsCl-based ultracentrifugation is used to form a collected band of DNA molecules, and E. coli remains an important experimental organism. More significantly, the simplicity and elegance of the experimental design make their experiment a timeless classic.