All sections Interactive
00Interactive

Three Ways to Change a Gene

The same stretch of DNA, edited three ways. The differences between these tools are the whole story of why some therapies are close and others are not.

An edited base pair
Human DNA · PCSK9 fragment Unedited

Pick a tool

Each one changes the same target differently. Watch what happens to the highlighted letters.

Evidence status Approved In trials Early research

The word "CRISPR" gets used as though it named one thing. It does not. There are now several distinct ways to change a gene, and the differences between them are not academic. They determine which diseases are treatable, how safe a therapy is, and how close it is to reaching patients. The tool above uses a fragment of PCSK9, a real gene that controls how much cholesterol the liver clears from the blood, and a real target of therapies in trials today.

Cutting is not the same as fixing

The original CRISPR-Cas9 system is best understood as molecular scissors guided to a specific address. It finds a chosen sequence and cuts straight through both strands of the double helix.1 What it does not do is repair anything. The cell notices the break and rushes to reseal it, and that emergency repair is usually sloppy, frequently deleting or inserting a few letters at the join.

For many therapies that sloppiness is the point. If the goal is to switch a gene off, a scrambled repair does the job perfectly well. This is roughly the strategy behind the approved sickle cell therapies Approved, which disable a regulatory element so the body produces more fetal hemoglobin. But if the goal is to correct a specific misspelling rather than break something, a double-strand cut is a blunt instrument, and every break carries some risk of rearrangement at the wrong site.

Most disease-causing mutations are single wrong letters. Cutting the page in half is a strange way to fix a typo.

Editing a single letter

Base editing solves that problem by not cutting at all. A base editor attaches a chemical enzyme to a deactivated version of the Cas protein. The Cas part still finds the right address, but instead of cutting, the chemical tool converts one DNA letter into another, most often A into G or C into T.2 The backbone of the helix is never severed.

This is a much gentler operation, and it is why base editing reached the clinic quickly for targets that need exactly this kind of change. The trial that switched off PCSK9 in the liver, lowering cholesterol substantially after a single infusion, used a base editor. In trials5 The limitation is in the name: it changes letters. It cannot insert missing sequence or remove extra sequence, and it can only make certain letter swaps.

Search and replace

Prime editing is the most flexible of the three. It carries its own template, a strip of genetic code spelling out the replacement, and writes that new sequence straight into the target using an enzyme called reverse transcriptase, which builds DNA from an RNA template.3 It nicks only one strand rather than cutting both. In principle it can perform any small substitution, insertion, or deletion, which covers the large majority of known disease-causing mutations.

The tradeoff is complexity. A prime editing complex is a larger machine to build and to deliver into cells, and the technology is younger, so there is less human safety data. The first clinical results are only now appearing, including a reported use in chronic granulomatous disease, a rare inherited disorder in which immune cells cannot properly kill bacteria, where the patient's cells were edited outside the body and then returned. Early research

Why the differences matter

A therapy is not chosen by which tool is newest. It is chosen by what the disease requires. Switching a gene off favors the cut. Correcting a single known misspelling favors base editing. Repairing a mutation that needs missing sequence restored requires prime editing or nothing at all. And in every case the harder problem is delivery, which is to say getting the machinery into the right cells in the right organ without provoking the immune system.

ToolWhat it doesMain strengthMain limit
CRISPR-Cas9 Cuts both DNA strands at a chosen address Efficient at disabling a gene Repair is error-prone; double-strand breaks carry rearrangement risk
Base editing Chemically converts one letter into another No double-strand break, so a gentler edit Only certain letter swaps; cannot add or remove sequence
Prime editing Writes a new sequence from its own template Handles substitutions, insertions, and deletions Larger and harder to deliver; least human data so far
Key Takeaways
  • CRISPR-Cas9 cuts DNA but does not repair it, and the cell's own emergency repair is what introduces changes.
  • Base editing converts one letter into another without severing the helix, which is why it reached patients quickly for suitable targets.
  • Prime editing can write new sequence and is the most flexible, but it is bigger, harder to deliver, and newer.
  • The choice of tool follows the disease, and delivery into the right cells remains the harder problem for all three.
Further Reading & References
  1. Jinek M, et al. A Programmable Dual-RNA-Guided DNA Endonuclease in Adaptive Bacterial Immunity. Science. 2012.
  2. Komor AC, et al. Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage. Nature. 2016.
  3. Anzalone AV, et al. Search-and-replace genome editing without double-strand breaks or donor DNA. Nature. 2019.
  4. Frangoul H, et al. Exagamglogene Autotemcel for Severe Sickle Cell Disease. New England Journal of Medicine. 2024.
Recent Developments · 2025 to 2026
  1. In Vivo Base Editing of PCSK9 with VERVE-102 for Hypercholesterolemia. New England Journal of Medicine. 2026. The base-editing trial that used the gene modelled in the tool above.