Replication

Watson and Crick saw at once how DNA could copy itself: the two strands separate, and each acts as a template for a new partner strand, built by base pairing. Each new DNA molecule then has one old strand and one new strand. This is semiconservative replication.

Meselson and Stahl (1958) proved it. They grew E. coli for many generations in a medium with ¹⁵N, a heavy (but not radioactive) form of nitrogen, so all their DNA was heavy. Then they moved the bacteria to ordinary ¹⁴N and took samples after each division. They separated DNA by density, spinning it in a caesium chloride (CsCl) gradient:

• After one generation, all the DNA was of intermediate density (hybrid): one heavy strand and one light.
• After two generations, half was hybrid and half was light.

That is exactly what semiconservative replication predicts. Taylor later showed the same in bean plants, using radioactive thymidine.

The machinery. The main enzyme is DNA-dependent DNA polymerase. It is very fast (about 2000 base pairs per second in E. coli) and very accurate. Its building blocks are deoxynucleoside triphosphates, which supply both the nucleotides and the energy for joining them. Unwinding the long helix all at once is impossible, so it opens a little at a time, forming a Y-shaped replication fork.

DNA polymerase can only add nucleotides in the 5'→3' direction. Because the two strands are anti-parallel, this causes a problem at the fork:

• On one template, the new strand grows continuously, following the fork.
• On the other, it must be made discontinuously, in short pieces going backwards, which are later joined by DNA ligase.

Replication starts at specific places called the origin of replication. In eukaryotes, it happens in the S-phase of the cell cycle; it must be finished before the cell divides.