Biology › Nucleic acids, genomes and protein synthesis › DNA replication: one old strand in every new molecule

DNA replication: one old strand in every new molecule

Every time a cell divides it must produce two complete copies of about 3.1 billion base pairs, and get almost all of them right. The mechanism that manages it keeps one original strand in each new molecule, and the experiment that proved it is one of the cleanest in biology.

Before this DNA structure and complementary base pairing · Enzymes and the induced fit model · ATP as an energy currency

What you should be able to do

• Explain what semi-conservative replication means, in terms of what happens to the original strands.
• Describe the roles of helicase and DNA polymerase, naming the bonds each one deals with.
• Explain why one new strand is made continuously and the other in fragments.
• Describe the Meselson-Stahl experiment and explain how its results eliminated two competing models.
• Explain how a replication error becomes a mutation, and why most cause no harm.

What 'semi-conservative' actually claims

Before the mechanism was known, three models were on the table. The conservative model said the original double helix stayed intact and an entirely new one was built alongside it. The dispersive model said the new molecules were patchworks, with stretches of old and new sequence scattered along both strands. The semi-conservative model said the two strands separate, each acts as a template, and each new molecule contains one original strand and one new one.

The word semi-conservative is doing exact work: half of each daughter molecule is conserved from the parent. Say that in an answer and you have the mark. Say 'half the DNA is conserved' and you have said something vaguer and, after several generations, untrue.

Semi-conservative replication

Replication in which the two strands of the parent molecule separate and each acts as a template, so every daughter molecule contains one original strand and one newly synthesised strand.

Template strand

A strand of DNA whose base sequence determines the sequence of the new strand built against it.

Unzipping, then building

Replication begins when DNA helicase travels along the molecule and breaks the hydrogen bonds between paired bases. The two strands separate and the helix unwinds, exposing the bases. Notice which bonds are broken: hydrogen bonds, across the molecule. The phosphodiester bonds along each backbone are covalent and are left alone, which is exactly why the strands come apart as two intact strands rather than as debris.

Free DNA nucleotides, already present in the nucleoplasm, then align against the exposed bases by complementary base pairing. An adenine on the template attracts a thymine nucleotide and nothing else; a cytosine attracts a guanine. The specificity of the pairing is what makes the copy accurate, and the accuracy is not a separate mechanism bolted on — it falls out of the chemistry of the bases.

DNA polymerase then catalyses the condensation reactions that join the aligned nucleotides into a continuous strand, forming a phosphodiester bond between each one and the next. The free nucleotides arrive as nucleoside triphosphates, so the energy for each bond comes from the two extra phosphates being released as the bond forms.

DNA polymerase can only build a strand in the 5′ to 3′ direction, because it can only attach a new nucleotide to a free 3′ hydroxyl group. The two template strands are antiparallel, so the enzyme cannot travel the same way along both. On one template it moves in the direction the fork is opening, and produces one long unbroken leading strand. On the other it must work away from the fork, so it waits for a stretch of template to be exposed, copies backwards along it, then jumps forward and does it again. The result is the lagging strand: a series of short pieces called Okazaki fragments, typically 100 to 200 nucleotides long in eukaryotes, which DNA ligase later joins into one continuous strand.

The commonest confusion here is thinking the lagging strand is built 3′ to 5′. It is not. Both new strands are built 5′ to 3′ without exception. What differs is the direction of travel relative to the fork, and that is what forces one of them into fragments.

Meselson and Stahl: how the argument was settled

In 1958 Matthew Meselson and Franklin Stahl found a way to tell the three models apart by weight. They grew E. coli for many generations in a medium whose only nitrogen source was the heavy isotope ¹⁵N. Nitrogen is in every base, so the bacteria built DNA that was measurably denser than normal. They then transferred the culture to a medium containing only ordinary ¹⁴N, so every nucleotide made from that point on was light, and sampled the DNA after each round of division.

Spinning the extracted DNA in a caesium chloride solution at very high speed sets up a density gradient, and each DNA molecule settles at the depth matching its own density. Heavy DNA sinks furthest. Light DNA stays highest. A hybrid molecule, one heavy strand and one light, settles halfway between.

After one generation there was a single band, and it sat exactly halfway between the heavy and light positions. The conservative model dies here. It predicts the original heavy molecule survives untouched and an entirely light molecule is built beside it, giving two bands — one heavy, one light — and no intermediate band at all. A conservative mechanism cannot produce a molecule of mixed density, because it never puts old and new material in the same molecule. One intermediate band and no heavy band rules it out outright.

That result alone does not distinguish semi-conservative from dispersive, because a patchwork molecule of half-old, half-new sequence would also be of intermediate density. The second generation separates them. The semi-conservative model predicts two bands: half the molecules still contain one heavy strand and stay intermediate, while half are now entirely light. The dispersive model predicts one band again, this time three quarters of the way toward light, because every molecule is still an even mixture. Two distinct bands appeared. Dispersive replication was finished.

When the copying goes wrong

Human DNA polymerase makes roughly one error per billion base pairs copied, after its own proofreading has corrected most of what it got wrong. That is remarkable accuracy, and across 3.1 billion base pairs it still means a handful of new errors in every cell division.

An uncorrected error is a mutation: a change in the base sequence of DNA. A base substitution swaps one base for another. An insertion or deletion adds or removes one, which shifts the reading of everything downstream and is usually far more damaging. Mutations also arise from mutagens — ultraviolet light, some chemicals in tobacco smoke, ionising radiation — rather than from replication alone.

Mutation

A change in the base sequence of DNA.

Substitution

A mutation in which one base is replaced by another.

Deletion

A mutation in which a base is lost, shifting the reading of every triplet that follows.

Most mutations have no effect. Some fall in non-coding stretches of DNA. Some change a triplet into another triplet coding for the same amino acid, which the next lesson explains. Some change an amino acid in a region of the protein where the substitution makes no difference to the folded shape. A minority land somewhere that matters, and of those a few are useful, which is where the variation that natural selection acts on comes from.

The classic example is worth quoting precisely: in sickle-cell anaemia a single base substitution in the gene for the β-globin chain of haemoglobin changes one triplet, one amino acid changes from glutamic acid to valine, and the altered haemoglobin polymerises at low oxygen concentrations and distorts the red blood cell. The gene sits on chromosome 11, so that is one base out of about 135 million on that chromosome.

In the exam

• Semi-conservative means each new molecule has one original strand and one new strand. Answers saying 'half the DNA is old' are too loose to score.
• Name the bonds separately. Helicase breaks hydrogen bonds between bases; DNA polymerase forms phosphodiester bonds along the backbone. Swapping them is a common and expensive error.
• Both new strands are built 5′ to 3′. The lagging strand is discontinuous because of the direction it must travel relative to the fork, not because the polymerase reverses.
• For Meselson-Stahl, quote what each model predicts before saying what was observed. The marks are for the elimination, not for describing the tubes.
• Give the isotopes the right way round: ¹⁵N is the heavy one the bacteria were grown in first, ¹⁴N the light one they were moved to.

Questions

Written to the command words the boards use. Try them on paper before opening a scheme: the marks go to points made, not to length.

← DNA and RNA: four letters and a rule about pairing · The genetic code, and getting a gene out of the nucleus →