Beyond CRISPR-Cas9: What Base Editing and Prime Editing Actually Enable

CRISPR-Cas9 gets all the headlines, but if you're paying attention to what's actually happening in genetic medicine right now, the more interesting work is happening with its successors. Base editing and prime editing don't work the way CRISPR does — and that difference matters enormously for what's treatable and what isn't.

Here's the problem with classical CRISPR-Cas9: it's a very good scissors. It cuts both strands of DNA at a precise location, which is genuinely remarkable. But what happens after the cut is largely out of your control. The cell's own repair machinery kicks in, and it has two main options. The messy one — non-homologous end joining — essentially glues the ends back together, often introducing small insertions or deletions that disrupt the gene. Useful if you want to knock a gene out. Less useful if you want to make a precise, specific change.

The clean option — homology-directed repair — can incorporate a template you provide, giving you actual control over the edit. But this pathway is unreliable, especially in cells that aren't actively dividing. In neurons, for example, where most inherited neurological diseases manifest, homology-directed repair is frustratingly inefficient.

## So what does base editing actually do?

Base editing, developed primarily by David Liu's lab at the Broad Institute starting around 2016, takes a different approach. Instead of cutting DNA, a base editor converts one DNA letter into another directly. Think of it less like scissors and more like a chemical pen with Wite-Out.

The two main classes convert C→T or A→G (and their complements). That covers a huge range of disease-causing point mutations — single-letter changes in the genetic code that account for roughly half of all known pathogenic variants. Sickle cell disease, for instance, is caused by a single A→T mutation in the beta-globin gene. A base editor can reverse that.

**The key advantage:** no double-strand break. No unpredictable repair machinery. The change happens enzymatically, at the specific position, without cutting through both DNA strands. This dramatically reduces the risk of off-target genomic instability.

> 🔬 **Quick note on scale:** There are roughly 15,000 known point mutations that cause human disease. Base editing is relevant to a large fraction of them — specifically those involving the four base conversions current editors can make.

## And prime editing goes further still

Prime editing — announced by Liu's lab in 2019 — is genuinely different again. If base editing is a targeted chemical modification, prime editing is closer to a find-and-replace function for DNA sequences.

A prime editor uses a modified Cas9 (one that nicks only one strand rather than cutting both) attached to a reverse transcriptase enzyme and a "pegRNA" — a specially designed guide RNA that also encodes the desired edit. The reverse transcriptase writes the new sequence directly into the nicked strand, using the pegRNA as a template. The cell then resolves the resulting mismatch, ideally incorporating the edit.

What this enables that base editing doesn't: insertions of new sequences (not just conversions), deletions of specific sequences, and all twelve possible base conversions rather than just four. Prime editing can, in principle, correct the vast majority of the 75,000 or so variants in the ClinVar database of clinically relevant genetic variants.

## But "in principle" is doing a lot of work there

Here's the weird part: in cell culture, prime editing works beautifully. In living organisms — especially in tissues that are hard to deliver large molecules into — efficiency drops. The pegRNA is larger and more structurally complex than a standard guide RNA. Delivery to the relevant tissue remains a serious engineering challenge.

The viral vectors (often AAV — adeno-associated virus) that deliver gene editing machinery have size limits, and some prime editor constructs are close to or exceed those limits. Split-intein systems that reassemble the editor inside the cell have helped, but they add complexity. This is a solvable problem, but it's not solved.

Prime editing is also slower to get things wrong — which is good! — but also slower to act. For acute conditions where you need rapid gene correction, that matters.

## Why this is the actual frontier

CRISPR-Cas9 got us from "gene editing is possible" to "gene editing is practical." Base editing and prime editing are getting us from "we can disrupt genes" to "we can write specific changes with precision." The approved sickle cell and beta-thalassemia therapies from 2023 used classical CRISPR to reactivate fetal hemoglobin — a clever workaround, not a direct correction. The next generation will likely fix the underlying mutation directly.

*I think that distinction — between workarounds and actual corrections — is what makes these tools so significant.* The question isn't whether base editing and prime editing will reach the clinic. They already have, in early trials. The question is how quickly the delivery problems get solved, and how many of those 75,000 disease variants become treatable once they do.

Beyond CRISPR-Cas9: What Base Editing and Prime Editing Actually Enable

// COMMENTS

ON THIS PAGE