Meselson–Stahl experiment
A 1957-1958 experiment with bacteria that showed DNA copies itself semiconservatively: each new molecule keeps one original strand and builds a new one.
- Year
- 1958
- Section
- Biology
History
The Meselson-Stahl experiment is a laboratory test with bacteria that showed how DNA, the molecule holding the instructions of living things, is copied. It mattered because it answered a question biologists had been asking for years: when a cell divides, how does it share its DNA between the two daughter cells? The work was carried out around 1957 and published in 1958, and its clever yet simple design made it one of the most cited experiments in biology.
To understand the problem, it helps to recall the shape of DNA. It is built from two strands twisted into a double helix and joined by pairs of chemical bases. After James Watson and Francis Crick proposed this structure, it was clear that the strands could separate and act as templates for new ones, but it was unclear what happened to the original strands. Three hypotheses competed. The semiconservative one, backed by Watson and Crick, held that each new molecule would keep one old strand and build one new strand. The conservative one held that the original molecule stayed intact while the copy was made entirely from scratch. The dispersive one, illustrated by a model proposed by Max Delbruck, pictured the DNA being broken into pieces, with old and new material mixed within both strands.
Meselson and Stahl needed a way to tell old DNA from new. Their solution relied on nitrogen, an element found in DNA's bases. Most nitrogen is the lighter 14N, but there is a heavier isotope, 15N, which behaves chemically the same with one slightly heavier atom. It is not radioactive, so it posed no danger, and DNA containing it still works normally. The trick was that this small difference in weight could be measured in a lab.
They grew the bacterium Escherichia coli for several generations in a medium containing 15N, so that its DNA became labelled with the heavy isotope. They extracted the DNA and spun it in a cesium chloride solution, a salt that forms a density gradient. During centrifugation, each molecule settles where its density matches that of the surrounding solution. DNA containing 15N settled lower than DNA from bacteria grown on 14N, confirming that the two could be told apart.
Next, they moved the labelled bacteria to a 14N medium and let them divide once. The resulting DNA was neither heavy nor light but had an intermediate density. This ruled out the conservative hypothesis, which would have produced separate heavy and light molecules and no intermediate ones. However, the result still fit both the semiconservative and dispersive models, since in both cases the DNA would show an average density.
To separate those two options, the team waited for a second division. The bacteria were allowed to replicate twice in 14N, and the DNA was analysed again. Two densities now appeared in equal amounts: one intermediate, matching the first generation, and one light, made only of 14N. The dispersive model would have produced a single density, lighter than the intermediate one, because the old DNA would have been spread evenly across all strands. The semiconservative model fit the data: half the molecules kept one original 15N strand paired with one new 14N strand, while the other half were built entirely from new material.
Over time the work became a reference point in molecular biology. Its design, which combined a clear question, a simple label and a precise measurement, is still taught as a model of good experimental reasoning. Semiconservative replication also explains why a cell can copy its genetic instructions faithfully and why occasional copying errors can be passed on and accumulate. Today, understanding this mechanism supports genetic research, molecular diagnosis and the development of treatments, including some for cancer, a disease in which faults in DNA copying play an important role. Density centrifugation and isotope labelling remain everyday tools in many laboratories.
Timeline
- 1957Meselson and Stahl carry out the experiment with E. coli bacteria labelled with 15N.
- 1958The results are published, confirming that DNA replication is semiconservative.
- Watson and Crick propose that DNA replication is semiconservative, the hypothesis this experiment would support.
- Max Delbruck proposes a dispersive model, one of the three hypotheses tested.
Trivia and anecdotes
A much-admired experiment
The work has been called the most beautiful experiment in biology, a description that reflects the elegance of its design.
A heavy but harmless isotope
15N has more mass than common nitrogen but is not radioactive, so it could serve as a label without risk to the bacteria or the staff.