replication fork

Take a zipper that is partway undone. At the open end the two halves splay apart into a Y shape, while below the slider they are still joined. As you keep unzipping, that Y-shaped junction travels down the zipper. The replication fork is exactly this travelling Y in DNA: the moving point where the double helix is being split into two single strands so that each can be copied.

At the fork, an enzyme called helicase runs ahead and unwinds the double helix, peeling the two strands apart. The freshly separated single strands are immediately coated and copied: a whole crew of proteins clusters at the fork to do the job, including the ones that lay down starter pieces, the main copying enzymes, and helpers that keep the strands from snapping back together or tangling. Because DNA usually opens at an internal origin, two forks normally form there and travel in opposite directions, each looking like a Y, together forming an expanding bubble of copied DNA.

The fork is where almost all the action of replication happens, so it is the natural place to understand the whole process. It also reveals a built-in complication: the two template strands run in opposite directions (they are antiparallel), yet the copying enzyme can only build a new strand in one direction. This forces one new strand to be made smoothly and the other in short backstitched pieces — the leading and lagging strand problem. A common misconception is that the fork is a fixed gate that DNA passes through; really it is a moving worksite that glides along the DNA as copying proceeds.