There are many existing hypotheses for the mechanism of DNA replication. There are three main models. The conservative model states that the parent DNA molecule remains completely intact. New DNA molecules are made of completely new strands. The semiconservative model states that each new DNA has one new strand and one parent strand. The dispersive model states that fragments of the parent are randomly passed on.
Previous experiments used radioisotope labels to study DNA. In this experiment, Meselsohn and Stahl used differences in density instead. The density differences allowed them to see how nucleotides move through replication. They put DNA in a solution of concentrated cesium chloride. Then, they centrifuged the DNA until a solid band forms at a particular position. The location corresponds to buoyant density of the sample. The density of the DNA increases with increasing centrifugal force in a gradient. The researchers labeled DNA with heavy nitrogen (N15) and unlabeled (N14) nitrogen. Centrifugation allowed the detection of the different densities of nitrogen isotope in DNA.
Meselsohn and Stahl analyzes DNA replication. They attempt to disprove the conservative and dispersive models of replication.
Researchers grew Escherichia coli B in N15 for 14 generations (Figure 3). They kept track of the amount of bacteria with microscope counts and colony assays. Then, they changed the medium to N14. They took samples of bacteria at various points in time. They lysed the bacteria with 15% sodium dodecyl sulfate to isolate the DNA. They centrifuged the DNA, and they determined the density with a microdensitometer (Figure 2). They took ultraviolet absorption photographs of the bands. The buoyant density of the DNA varies has a direct relationship with the fraction of N15 it contains. Thus, researchers can determine the amount of N15 based on the position of the band. They can compare the band position to the labeled and unlabeled DNA bands. They also used heat denaturation to separate the two strands and analyze them.
The parent DNA molecule is completely labeled with N15. After one generation, the two daughter DNA molecules are both half-labeled. Both daughter DNA molecules contain equal amounts of labeled nitrogen. After two generations, half of the DNA is half-labeled and the other half is unlabeled (Figure 4). Thus, the daughter molecule receive equal amounts of nitrogen from the parent molecule. And, strands remain intact through many generations. Moreover, each daughter molecule receives one parental strand after replication. Otherwise, there would be a mix of completely labeled strands and unlabeled strands. After each replication, the number of molecules double.
When DNA is heated, its strands come apart because there is enough energy to break the hydrogen bonds. Single-stranded DNA is denser than the double helix. Still, the researchers were able to determine the DNA density from a microdensitometer. Meselsohn and Stahl also had a way of determining the concentration of DNA based on the band width. This ability still held even with single strands of DNA.
Next, Meselsohn and Stahl heat denatured hybrid DNA. The hybrid DNA (two strands) consisted of half N14 and half N15. But, they did not know the amount of each nitrogen in each strand. When heat denaturation separated the DNA strands, there were two distinct density peaks. One peak corresponded to denatured N14 and one corresponded to denatured N15. As a control, they tested the density gradient of a heated mixture of N14 DNA and N15 DNA. The density peaks resembled the peaks produced by the denatured hybrid DNA. Thus, heat did indeed separate the hybrid DNA. Furthermore, each DNA strand is conserved through replication. One of the new strands consists completely of N14 DNA. The other new strand consists of N15 DNA only (Figure 9). These results support the semi-conservative model of DNA replication.
The results are consistent with Watson and Crick's model of DNA replication. In the Watson and Crick model, the DNA strands first separate. The separation exposes the hydrogen bond sites of the bases. Because of base-pairing restrictions, each strand is a template for the daughter molecules. The template allows for the synthesis of a complementary strand. Thus, each daughter molecule has one parent strand and one newly synthesized strand (Figure 6). Still, this experiment does not prove the DNA structure proposed by Watson and Crick. These results support the semi-conservative model of DNA replication. They disprove the conservative and dispersive models.