How CRISPR edits a gene, and what base editing added to it

The problem CRISPR solves

Every cell carries DNA written in a four-letter code: A, C, G and T. Genes are stretches of that code that instruct the cell to build particular proteins. Change the code in the wrong place and a protein can stop working, work too well, or work incorrectly. Many inherited diseases come down to a single wrong letter, or a few, in a single gene.

For decades, deliberately altering DNA in a living cell was slow, expensive and imprecise. CRISPR changed that. It gave scientists a way to point at almost any spot in a genome and make a change there, using tools borrowed from bacteria, which use CRISPR systems as a defence against viruses.

How CRISPR-Cas9 actually works

The version of CRISPR most people mean when they say the word uses a protein called Cas9 alongside a short piece of RNA called a guide RNA.

The guide RNA is designed by scientists to match the exact DNA sequence they want to edit, letter for letter. It acts like a search term. Once introduced into a cell, the guide RNA pairs up with Cas9 and steers it through the genome until it finds the matching sequence.

When Cas9 locates the target, it does something quite blunt: it cuts both strands of the DNA double helix at that point, creating what’s called a double-strand break.

The cell then tries to repair the damage, and this is where the actual editing happens. There are two main repair routes:

• Non-homologous end joining (NHEJ) simply glues the cut ends back together, but it’s sloppy and often adds or deletes a few letters in the process. This is useful for disabling a gene entirely, since scrambling its code stops it working.
• Homology-directed repair (HDR) uses a template, a separate piece of DNA supplied by the scientist, to patch the break according to a precise new sequence. This route can insert or correct specific code, but it only happens efficiently in certain types of dividing cells.

This is CRISPR’s core trick: guide RNA for targeting, Cas9 for cutting, and the cell’s own repair machinery for making the change stick.

Why cutting DNA is a double-edged sword

Deliberately breaking both strands of DNA is powerful, but it’s also risky. Double-strand breaks are exactly the kind of damage cells evolved to fear, because getting the repair wrong can cause larger deletions, rearrangements, or edits in the wrong location entirely, known as off-target effects. HDR, the precise repair route, is also inefficient in many cell types, particularly ones that aren’t actively dividing, such as most cells in an adult body.

For a while, this meant CRISPR was excellent at knocking genes out but clumsy at making the kind of small, exact, single-letter correction that many genetic diseases actually require.

What base editing changed

Base editing, developed by adapting the CRISPR system rather than replacing it, tackles that problem by removing the cut entirely.

A base editor still uses a guide RNA to find the right spot in the genome, just like standard CRISPR. But instead of a fully active Cas9 that slices both DNA strands, it uses a modified Cas9 that only nicks one strand, or doesn’t cut at all. Fused to this modified Cas9 is a second protein, an enzyme called a deaminase, that performs a direct chemical conversion of one DNA letter into another, right at the targeted position.

In practice, this allows scientists to convert one specific base pair into another (for example, correcting a C where an A should be) without ever creating a double-strand break. Because the DNA backbone stays intact throughout, the cell doesn’t need to invoke its higher-risk repair pathways, which lowers the chance of the large unwanted deletions and rearrangements associated with standard CRISPR cutting.

The trade-off is scope. Base editing can only make certain types of letter-to-letter conversions, not any arbitrary change, and it can’t insert or remove larger chunks of DNA the way HDR-based CRISPR editing can. It is a scalpel for specific, well-understood single-letter mutations, not a general-purpose rewriting tool. A related technique called prime editing, developed afterwards, extends this same cut-free logic to a wider range of edits, including small insertions and deletions.

Why this matters beyond the lab

The distinction matters because so many known disease-causing mutations are single-letter changes. Reducing the risk of off-target damage while making that exact kind of correction is a meaningful step toward using gene editing directly in patients, not just in lab-grown cells. In the UK, human embryo and reproductive gene editing remains tightly regulated, and any clinical use of these tools in patients goes through dedicated regulatory approval processes. Agricultural gene editing in England is governed separately, under legislation that treats precision-bred organisms differently from older genetic modification rules, reflecting the same underlying idea: that precise, small edits carry a different risk profile to older techniques.

Understanding the difference between cutting DNA and directly converting a letter within it is the key to understanding why base editing has generated so much interest as a more surgical successor to the original CRISPR toolkit, rather than a replacement for it.