Three Models for DNA Replication

The original paper by Watson and Crick suggested that DNA replication was semi-conservative; during replication the original two strands separate, and each would function as the template for a new partner. Each molecule would therefore consist of one new and one old strand (Fig 1c). Two other models for replication were suggested. The conservative model (Fig 1a, below) proposed that the original double stranded helix would serve as a template for synthesizing new DNA, but after replication, the new strands would somehow separate and come together and the original strands would likewise remain together. The model of dispersive replication (Fig. 1b) proposed that fragments of the original molecule served as templates, synthesizing two new molecules, each of which contained old and new parts, perhaps at random.

Fig. 1. Three models for DNA Replication

Confirmation of semi-conservative replication came from experiments by Matthew Meselson and Franklin Stahl in 1957. They devised an experiment using an isotope of nitrogen (heavy nitrogen, ¹⁵N instead of the more common ¹⁴N) that allowed them to distinguish old DNA strands from new ones. They developed a centrifugation procedure (Fig. 2) that allowed them to determine the density of DNA from a single sample.

Fig. 2. Density Gradient Centrifugation. When a solution of cesium chloride is centrifuged at extremely high speed, the cesium ions tend to sink slightly, forming a density gradient along the tube. Another substance in the tube will float at the point where its density matches that of the gradient. In this illustration, the red DNA molecules aggregate in a single band.

Meselson and Stahl grew the bacterium Escherichia coli for 17 generations on a medium in which the only nitrogen source was the heavy nitrogen isotope, ¹⁵N. Eventually all the DNA in the bacteria was ¹⁵N. The researchers then changed the nitrogen source to the normal ¹⁴N. Cell samples were removed at intervals and DNA was extracted from the cells and subjected to centrifugation. They found that after the DNA was duplicated and the cells divided to produce each new generation, the DNA banding pattern was different from the original banding (Fig. 3). In the initial generation, the DNA was uniformly labeled with ¹⁵N and therefore relatively dense. After one generation, when the DNA had been duplicated once, all the DNA was of intermediate density. After two generations, there were two equally large DNA bands: one of low density and one of intermediate density. In samples from subsequent generations, the proportion of low-density DNA increased steadily. This provided unequivocal support for the semi-conservative model. What would the banding pattern have looked like with the conservative model? the dispersive model?

Fig. 3. Meselson-Stahl experiment supports the semiconservative model of DNA replication. The two strands of each DNA molecule stay together during extraction and centrifugation. Before the bacteria reproduce the first time in the light medium, all DNA (parental) is heavy. If each strand served as a template for the second strand, DNA of the first generation would be of an intermediate density and half the DNA from the second generation would be intermediate and half light, as was observed.

Download Printable (PDF) Version