How CRISPR base editing differs from standard CRISPR cutting

Base editing differs from standard CRISPR by not cutting the DNA double helix at all. Standard CRISPR acts like scissors: it snips both strands and lets the cell's repair machinery patch the gap, imperfectly and unpredictably. Base editing, first described by David Liu's laboratory at Harvard University in a 2016 paper in Nature, chemically converts one DNA letter into another in place. Prime editing, its younger sibling from the same group in 2019, writes short new sequences without double-strand breaks.

Engevity News publishes information about research, not medical advice. Gene-editing therapies remain experimental in most settings, and treatment decisions belong with specialists.

What does standard CRISPR actually do to a cell?

The familiar CRISPR system has two parts: a guide RNA that locates a chosen stretch of DNA by matching its sequence, and an enzyme, Cas9, that cuts both strands there. What happens next depends on the cell. The most common repair pathway, called non-homologous end joining, glues the ends back together quickly and sloppily, often adding or deleting a few letters. That imprecision is a feature when the goal is to knock out a gene, and a bug when the goal is to correct a mutation.

Most approved CRISPR medicines exploit the sloppy pathway. Casgevy, a treatment for sickle cell disease that regulators in the United Kingdom and the United States approved in late 2023, does not repair the sickle mutation; it disrupts a different gene so that patients produce fetal hemoglobin instead.

How can you edit DNA without cutting it?

Base editors take the CRISPR scalpel and blunt it. Liu's team fused a crippled Cas9 — one that nicks only one strand, or none — with enzymes from nature's own repair toolbox. One type of base editor chemically converts the letter C into T; another converts A into G. The guide RNA parks the editor at the right spot, the attached enzyme performs the chemical swap on a single letter, and the cell's ordinary replication copies the change into the paired strand.

The limitation is arithmetic: there are four DNA letters, and each editor handles only one conversion. A C-to-T editor cannot make any other change. It is a pen that writes exactly one character — powerful for the thousands of diseases caused by single-letter mutations, useless for the rest.

What does prime editing add to the toolbox?

Prime editing, described in a 2019 Nature paper from the same laboratory, extends the reach. Its editor nicks one strand and uses an extended guide RNA as a template to write a short, chosen sequence into the gap — up to dozens of letters — again without severing the double helix. In principle, this covers the majority of known pathogenic mutations. In practice, laboratories have found that efficiency varies widely from site to site, and the system is larger and harder to deliver into cells than base editors.

Why do the differences matter medically?

The double-strand break is standard CRISPR's central risk. Cuts can trigger large unintended deletions, rearrangements of chromosomes, and a cellular alarm response; off-target cuts elsewhere in the genome are a separate hazard. Avoiding the break does not eliminate off-target chemistry, but it removes one class of damage entirely — which is why base editing was quickly adopted where cutting is most dangerous, such as cells that will be returned to patients.

The technology has already reached people. In 2022, a base-edited therapy developed at University College London treated an infant with drug-resistant leukemia in what the team described as the first use of base-edited cells in a person; the case was published in 2023. Since then, several clinical trials of base-edited cancer and immune cells have opened in the United States and Europe. These are small, early trials; the field's published experience remains a handful of patients.

How do we know the edits are safe?

Laboratories test editors in cultured cells and animals, sequencing the whole genome to hunt for unintended changes, and regulators require such data before human trials. But the honest limitations are well documented. Sequencing can miss certain kinds of rearrangements; long-term follow-up for any gene-editing therapy — standard or base-edited — is measured in years and is still accumulating; and the first treated patients are, inevitably, the experiment. Liu himself has repeatedly cautioned in public lectures and reviews that delivery, not editing chemistry, is now the field's bottleneck: getting a large editing machine into the right cells, in the right tissue, in the body, remains hard.

What about editing inside the body?

Most editing to date happens outside the patient: cells are edited in a laboratory and infused back. Editing inside a living body — in the liver, the eye, or the bone marrow — is harder, because the editing machinery must be delivered through the bloodstream and reach enough of the right cells. Standard CRISPR led the way: in 2021, Intellia and partners dosed the first systemic in-body CRISPR therapy in trials for a rare liver disease, with interim results reported through 2023 showing reduced pathogenic protein. Base editing followed into similar territory, with lipid-particle delivery programs entering clinical trials in 2023 and 2024. Results so far are measured in protein biomarkers and safety tables from small cohorts, not cures. In-body editing, whatever the editor, remains the youngest and most closely watched branch of the field.

Which approach will win?

Probably none outright. Scissors, pens, and template writers each fit different diseases, and clinicians increasingly speak of a toolbox rather than a competition. The sickle cell approval shows cutting is good enough for some goals; base editing shows precision pays where cuts are risky. The next decade of trials, not press releases, will assign the roles.