Evidence for semi-conservative replication, and telomeres

25 min

Testing a beautiful idea

Watson and Crick suggested in 1953 that each strand of DNA serves as a template for a new partner, so that every copy keeps one old strand. This is semi-conservative replication. It was an attractive idea, but other schemes were possible, and an idea is not evidence. In 1958 Matthew Meselson and Franklin Stahl, at the California Institute of Technology, designed an experiment to tell the possibilities apart. It is often called the most beautiful experiment in biology. This lesson also looks at a problem the copying machinery cannot solve at the ends of chromosomes, and how cells deal with it.

Three hypotheses

Conservative: the parent molecule stays intact, and an entirely new molecule is made.
Semi-conservative: each new molecule has one parental strand and one new strand.
Dispersive: both strands of every new molecule are patchworks of old and new DNA.

Definition

Isotope: an atom of the same element with a different number of neutrons. Nitrogen-15 (¹⁵N) is heavier than ordinary nitrogen-14 (¹⁴N); both are built into the bases of DNA.

Density-gradient centrifugation: spinning a sample at very high speed in a solution, here caesium chloride, that becomes denser towards the bottom of the tube. Each molecule settles at the level where the solution matches its own density, forming a band.

The Meselson–Stahl experiment

The bacterium E. coli was grown for many generations in a medium whose only nitrogen source was ¹⁵NH₄Cl, so all its DNA was "heavy" (¹⁵N/¹⁵N).
The bacteria were transferred to a medium containing only ordinary ¹⁴N. From then on, all newly made DNA strands were "light".
Samples were taken at the start and after each generation (each round of replication). DNA was extracted and centrifuged in caesium chloride, and the bands were photographed using ultraviolet light.

Interpreting the bands

Generation 1: one intermediate band. Every molecule was a hybrid, half ¹⁵N and half ¹⁴N. The conservative model predicted two bands, heavy and light, so it was rejected.
Generation 2: two equal bands, intermediate and light. The dispersive model predicted a single band, a little lighter than intermediate, so it was rejected. The semi-conservative model predicted exactly this: two hybrid molecules and two light ones.
Further evidence: when Meselson and Stahl heated generation-1 DNA to separate its strands, half the single strands were heavy and half were light, as semi-conservative replication requires.

Worked example 1: predicting later generations

Using the semi-conservative model, predict the percentages of hybrid and light DNA molecules after (a) 3 generations, (b) 4 generations in ¹⁴N.

Start with 1 heavy molecule. Its two heavy strands are never broken up, so however many rounds occur, exactly 2 molecules contain a heavy strand, and these are hybrids.
After $n$ rounds there are $2^{n}$ molecules, so the fraction that are hybrid is $\frac{2}{2^n}$ .
(a) $n = 3$ : $\frac{2}{8} = 25\,\%$ hybrid, 75 % light. (b) $n = 4$ : $\frac{2}{16} = 12.5\,\%$ hybrid, 87.5 % light.

Check against Figure 2: in generation 3 the light band is three times as thick as the hybrid band (75 : 25).

Definition

Telomere: a region of repeated, non-coding DNA at each end of a linear chromosome; in humans and other vertebrates the repeated sequence is TTAGGG.

Telomerase: an enzyme that adds telomere repeats to the ends of chromosomes, using an RNA template that is part of the enzyme.

Telomeres and continued replication

Bacteria have circular chromosomes, which have no ends. Eukaryotic chromosomes are linear, and their ends cause a problem, the end-replication problem. On the lagging strand, the last Okazaki fragment begins with an RNA primer at the very tip of the chromosome. When that primer is removed, DNA polymerase cannot fill the gap, because it can only add nucleotides to an existing 3′ end and there is none beyond the gap. So each new strand is a little shorter than its template, and the chromosome loses some DNA from its ends at every replication.

Telomeres solve the problem in two ways. They are long runs of a short repeat, several thousand base pairs long in humans, so what is lost at each division is repeat DNA, not genes. And with special proteins they form a cap that stops the cell's repair systems from treating the chromosome ends as broken DNA and joining chromosomes together.

In most human body cells telomeres shorten with every division. In 1961 Leonard Hayflick found that normal human cells grown in culture stop dividing after about 40 to 60 divisions. Short telomeres are one cause of this limit: when the telomeres become too short, the cell stops dividing. Some cells must go on dividing for a lifetime: germ cells that make sperm and eggs, and stem cells. They use telomerase to rebuild their telomeres. Most cancer cells also switch telomerase on, which lets them divide without limit. Elizabeth Blackburn, Carol Greider and Jack Szostak shared the 2009 Nobel Prize in Physiology or Medicine for discovering how telomeres and telomerase protect chromosomes.

Worked example 2: how many divisions?

Illustrative figures: a cell's shortest telomere is 9 000 base pairs long and loses 120 base pairs at each division. The cell stops dividing when the telomere falls to 3 000 base pairs. How many more divisions can it make?

DNA that can be lost: $9\,000 - 3\,000 = 6\,000$ base pairs.
Divisions: $\frac{6\,000}{120} = 50$ .

This is close to the Hayflick limit. With telomerase active, the telomere is rebuilt after each division, and the count never runs down.

Bozeman Science: "Meselson-Stahl Experiment". Watch for how each of the three hypotheses predicts different bands, and which result rules out each wrong model.

Warning

Common mistakes

Saying that generation 1 alone proves semi-conservative replication. It rules out only the conservative model; the dispersive model also predicts one intermediate band. You need generation 2.
"The bands are made of ¹⁵N." The bands are DNA molecules; their density depends on how much ¹⁵N they contain.
"Telomeres are genes for ageing." Telomeres are non-coding repeats; their shortening is one of many factors linked to ageing.

Tip

Answering Meselson–Stahl questions: draw the molecules strand by strand for each generation, colouring old strands and new strands differently, then sort the molecules into heavy, hybrid and light. The bands follow directly.

✦

Practice: the evidence and telomeres

Describe the three hypotheses for DNA replication.
Explain why Meselson and Stahl grew the bacteria in ¹⁵N for many generations before the experiment.
Which model was ruled out by the generation-1 result, and why?
Which model was ruled out by the generation-2 result, and why?
Predict the bands that would appear after one generation in ¹⁴N if replication were conservative.
Bacteria grown in ¹⁴N are transferred to ¹⁵N. Using the semi-conservative model, predict the percentages of light, hybrid and heavy DNA after two generations.
Explain why linear chromosomes get shorter at each replication.
Evaluate the claim that drugs that switch telomerase on in all body cells would be a safe way to slow ageing.

Show answersHide answers

Conservative: the parent molecule stays whole and a wholly new molecule is made. Semi-conservative: each new molecule has one old and one new strand. Dispersive: both strands of each new molecule are mixtures of old and new pieces.
So that all the DNA, both strands, was labelled with ¹⁵N (heavy) at the start, giving a single known heavy band.
Conservative: it predicts one heavy and one light band, but only one intermediate band was seen.
Dispersive: it predicts one band of intermediate-to-light density, but two separate bands, intermediate and light, were seen.
Two bands: one heavy (the original molecule) and one light (the new molecule), in equal amounts.
After two rounds: 2 hybrid molecules (50 %) and 2 heavy molecules (50 %); 0 % light. (The original light strands stay in 2 hybrids; all other strands are heavy.)
On the lagging strand, the RNA primer at the very end of the chromosome is removed and DNA polymerase cannot fill the gap, because there is no 3′ end to extend from. Each new strand is therefore shorter than its template.
Not safe: telomerase lets cells divide without limit, and most cancers depend on it, so switching it on everywhere could raise the risk of cancer. Ageing also has many other causes, so the benefit is uncertain.

Key points

Meselson and Stahl (1958): ¹⁵N to ¹⁴N transfer and caesium chloride centrifugation.
Generation 1: one hybrid band (rules out conservative). Generation 2: hybrid and light bands (rules out dispersive).
After $n$ generations, $\frac{2}{2^n}$ of the molecules are hybrid.
Telomeres (TTAGGG repeats) buffer and cap chromosome ends; telomerase rebuilds them in germ cells, stem cells and most cancer cells.

1Testing a beautiful idea

2Three hypotheses

3The Meselson–Stahl experiment

Interpreting the bands

4Telomeres and continued replication

Testing a beautiful idea

Three hypotheses

The Meselson–Stahl experiment

Telomeres and continued replication