28. Fork Reversal (1,320)

Let's reset the scene. A replication fork has stalled. The polymerase cannot move forward due to an unrepaired lesion that it has reached. If the lesion isn't immediately repaired or bypassed by TLS or re-priming in the first roughly 20-30 minutes after RPA-single-stranded DNA detection, the fork risks collapsing.

The cell's solution to this is highly mechanical: it literally pushes the replication fork backward. This maneuver, called fork reversal, moves the lesion away from the stuck DNA polymerase and places it back in the context of double-stranded DNA where repair enzymes can access it more easily.

Fork reversal isn't the easiest cellular process to visualize, so let me use an analogy: a zipper.

During normal DNA replication, the CMG helicase moves along the double helix separating the two strands so that each one--both the leading- and lagging-template strands-- can be replicated. In our analogy, the CMG helicase is the zipper handle pulling the zipper open.

But if a polymerase stalls at a lesion, instead of trying to force the zipper open past the obstruction, the cell temporarily slides the zipper handle backward, allowing the teeth of the zipper behind it to close again. That backward motion is fork reversal.

As the fork moves backward, the two template strands re-anneal with each other, while the two newly synthesized strands also pair with each other, forming a four-way DNA junction called a chicken-foot structure. This isn't easy to visualize, but let's wait a bit before clarifying the picture. Best to go step-by-step through this extremely tricky maneuver.

RAD51 arrives

When a DNA polymerase stalls at a particularly repair- or lesion bypass-resistant lesion, specialized proteins (that I'll introduce shortly) assemble at the fork. Their job is to recruit one of the most amazing proteins in genome replication and repair: RAD51.

RAD51 is best known for its role in homologous recombination (HR), one of the most accurate and amazing repair processes in molecular biology (and the subject of the next chapter). RAD51 makes both HR and fork reversal possible because it has a unique and powerful ability: it can help one DNA strand find and align with the equivalent sequence on its sister chromatid (i.e., the other similar chromosome).

Most proteins interact with DNA based on shape and/or chemical charge. But RAD51 can search for and recognize two DNA strands that contain exactly same genetic information. It's like using a search function in an ebook to find a specific sentence somewhere in the book. This ability is truly extraordinary.

Loading RAD51

At a stalled fork, the proteins BRCA1, BRCA2, and PALB2 collaborate to load RAD51 proteins onto the DNA.

You may recognize the names BRCA1 and BRCA2. Mutations in thse genes dramatically increase the risk of breast and ovarian cancer. ("BRCA" is an acronym for "breast cancer.") Their importance stems from their normal role in protecting the genome during replication and repair.

In the context of replication, BRCA1 helps determine whether RAD51 activity is appropriate at the stalled fork. Once the decision is made, PALB2 connects BRCA2 to BRCA1. BRCA2 then delivers clusters of RAD51 proteins to the DNA where they assemble into a filament along the newly synthesized strands near the stalled fork. This filament performs two critical functions.

First, it stabilizes the fragile DNA structure created by the stalled fork. The RAD51 filament is much more protective of single-stranded DNA than RPA proteins.

Second, it controls the activity of a nuclease (an enzyme that chew up DNA) called MRE11. MRE11 chews back the newly synthesized leading strand by about 10-50 nucleotides. This controlled trimming releases the stalled DNA polymerase and effectively resets the fork.

Without RAD51, this trimming by MRE11 would become uncontrolled and thousands of nucleotides could be degraded, severely damaging the replication fork. So RAD51 thus acts as both protector of single-stranded DNA and regulator of MRE11.

Fork reversal

Once the fork has been stabilized and the DNA polymerase has been disengaged, fork reversal begins.

The first step is carried out by specialized motor proteins called translocases. Translocases are molecular transporters that convert chemical energy into directed movement of molecules along DNA, RNA, or across cellular membranes. These enzymes are going to literally push the replication fork backward along the DNA.

As the fork moves backward, the two original template strands re-anneal with each other, just like the teeth of a zipper coming back together. But what happens to the newly synthesized strands that were just built on those two template strands?

They detatch from their respective template strands and, being complementary to each other, pair with each other. The result is a four-way DNA junction known as a reversed fork or a "chicken-foot" structure.

RAD51 coats the paired nascent strands of this new arm, stabilizing the structure and preventing it from collapsing.

At this point the lesion, which originally blocked the polymerase, has been pushed back into a region of normal double-stranded DNA where repair enzymes can now reach it.

Repairing the DNA around the fork

Before replication can restart, the reversed fork requires some tidying up. Because MRE11 trimmed the leading strand, the two nascent strands are no longer the same length. In addition, the lagging strand often contains small gaps between Okazaki fragments.

These problems are corrected via repair synthesis, carried out primarily by DNA polymerase delta. This synthesis doesn't have to be extremely accurate. The goal is to restore physical continuity to the DNA strands so the fork can safely restart.

Restarting replication

Once repairs are complete, the reversed fork must be restored back to its normal structure. This is the reverse of fork reversal. The helicase RECQ1 drives this process by pushing the DNA junction forward, undoing the reversed structure and restoring the familiar three-way replication fork.

Replication can now resume. The zipper begins opening again!

Repairing DNA-Protein Crosslinks (DPCs)

One particularly challenging kind of lesion is a DNA-protein crosslink (DPC)--a protein that has become chemically bonded to DNA. These lesions are large and physically block DNA polymerases.

After fork reversal pushes the DPC back into double-stranded DNA, the cell uses a specialized protease called SPRTN.

SPRTN is active only during DNA replication. It recognizes stalled forks and then digests the crosslinked protein into small peptide fragments.

However, SPRTN cannot remove the final amino acid that remains chemically attached to the DNA. This small remnant--called a peptide-DNA adduct--is then removed by the nucleotide excision repair (NER) pathway.

Together SPRTN and NER form a two-step hybrid repair system for DNA-protein crosslinks.

This pathway is essential to humans. Mutations in SPRTN cause a rare disorder called Ruijs-Aalfs syndrome, characterized by chromosomal instability, premature aging, and early-onset cancer.

But when replication encounters serious obstacles, the response cannot remain confined to a single fork. Signals must spread outward in order to slow replication elsewhere in the genome and mobilize repair systems across the cell--a coordinated global response that we will examine in the next chapter.