24. Errors II: RNA nucleotides (1145; 7/30)

Surprisingly, the most common type of DNA polymerase error is not the kind of mismatch, insertion, or deletion we covered in the last chapter. It's the insertion of a RNA nucleotide into a growing chain instead of a DNA nucleotide.

When that occurs, the bases do pair correctly, but the small chemical difference between an RNA nucleotide and a DNA nucleotide alters the structure of the DNA backbone and makes it prone to breakage. RNA nucleotide incorporations must be fixed.

How often does this occur? Recall that the initial DNA nucleotide error rate (before polymerase proofreading) was one error every 10,000 to 100,000 nucleotide additions. In stark contrast, the RNA nucleotide insertion rate is roughly one every 1,000 to 2,000 nucleotides!

So RER is not cleaning up rare accidents. RER will fix millions of RNA nucleotide incorporations during one genome replication. In comparison, MMR probably corrects tens to hundred of errors that have escaped proofreading. Think of RER as continuous sanitation of the genome and MMR as a precision service called to repair a relatively small number of dangerous errors.

There are two reasons why RNA nucleotide incorporation is so common. First, DNA polymerase's ability to discriminate the two isn't perfect. DNA and RNA nucleotides differ only by one very small chemical attachment.

Also, though, RNA nucleotides are present in the nucleus at 20-100 times higher concentration than DNA nucleotides. This is because they're the raw materials (monomers) of mRNAs, which the cell is synthesizing constantly. Because they're ubiquitous, they're easily grabbed and inserted.

Given how frequently RNA nucleotides are inserted, the cell requires a very effective way to remove them. DNA polymerase proofreading corrects them somewhat, but much less efficiently than it corrects DNA nucleotide mis-incorporations. The main pathway for correcting them is called ribonucleotide excision repair (RER) (RNA nucleotides are formally called "ribonucleotides").

Ribonucleotide Excision Repair (RER)

Like MMR, RER is a cut-and-patch pathway, but with some differences. Because it's focused on removing an RNA nucleotide, RER uses many of the same enzymes that remove the RNA portions of the primers used to initiate Okazaki fragments in lagging strand synthesis.

Also, whereas the MMR pathway begins near the replisome with error recognition but is often completed behind it as the replication fork continues to move forward, all of the enzymes involved in RER have PIP-boxes and are PCNA-linked. This makes the entire pathway more replication-associated than MMR. But that's not a hard rule. RER can also occur after the fork has moved on.

It's also worth emphasizing here the dual role of the PCNA sliding clamp. Yes, it's essential for DNA polymerases to function optimally. But in RER and other situations, PCNA also serves as a mobile organizing platform, or toolbelt. Its three PIP-binding pockets help the right enzymes (those with PIP boxes) act in the right place and in the right order.

Being a cut-and-patch pathway, RER starts with a detector protein, but a different one: RNase H2. In fact, RNase H2 is a protein complex made of three different proteins, one of which tethers to PCNA. This detector, unlike the MutSα, performs both of the first two steps in the pathway.

First, the enzyme detects the RNA nucleotide, distinguishing it from a DNA nucleotide based on the extremely small chemical difference between them. But it's also an endonuclease that makes the initial cut, or nick, on the new DNA strand immediately in front of the RNA nucleotide.

In MMR, the next step after nicking involved an exonuclease chewing back the new strand. But RER doesn't involve an exonuclease. In RER, once the nick is made, a new PCNA sliding clamp and DNA polymerase arrive.

This polymerase synthesizes DNA starting at the nick and continues a short distance. As it does, it displaces the strand containing the RNA nucleotide and several of the DNA nucleotides that follow, creating a flap that's up to about 10 nucleotides long. The flap is then cut off by a specific cleaving enzyme--call it a a flap cutter. This is the same enzyme that performs the same task in lagging strand synthesis.

Once the flap has been cleaved off, only one more job remains: connecting the two abutting single strands. That falls to another familiar enzyme: DNA ligase. It's localized to the site via PIP-box association with PCNA. Once strand continuity is restored, the repair is complete.

Because RNA nucleotide mis-incorporations are so common, RER is an obligatory quality control step in a DNA manufacturing process. RER removes about 99-99.9% of mis-incorporated RNA nucleotides. The remaining ones are handled later by other repair mechanisms.

MMR and RER... One More Difference

MMR and RER correct fundamentally different kinds of errors. MMR corrects an error in information: a wrong, extra or missing letter. This error could potentially alter codon reading. RER corrects an error in material. The RNA nucleotide will be correct in terms of pairing, but its chemical structure will be wrong.

In terms of targeting the right nucleotide for repair, this has major implications. With RER, it's relatively easy for the cell to identify the problem RNA nucleotide since it's a fundamentally different molecule than a DNA nucleotide.

But what about MMR's correction of mismatches, insertions and deletions? In the case of a mismatch, for example, how does the cell know which of the two paired nucleotides is the incorrect one? How does it know which of the two strands to repair? The error itself doesn't contain enough information.

Thus, the MMR enzymes look for clues. These clues take two forms depending on whether the mismatch is on the leading strand or the lagging strand. Let's consider the lagging strand first.

One characteristic of lagging strand synthesis is the many nicks, or small gaps, on the newly synthesized strand between Okazaki fragments before DNA ligase arrives to connect them. The MMR pathway uses this as its clue: fix the strand with the nicks. This will be the new strand--the one with the error. Thus, repair must occur fairly rapidly or the nicks will be filled and thus absent.

On the leading strand there are no nicks since DNA synthesis is continuous. But there is still a clue. It turns out that the PCNA sliding clamp's front face is different than its back face. The cell uses this asymmetry to know which of the two strands is the new strand (i.e., the strand that enters PCNA back face with respect to the direction of synthesis 5' end first). Tricky, don't you think?

With that we wrap up the repair of DNA polymerase errors. But polymerase errors are only one source of genomic insult that the cell must deal with. The others are lesions. These are chemical injuries to DNA that arise independently of replication and that can be far more disruptive.

We turn next to lesions and some even more remarkable repair systems that keep them in check.