Efficient gene insertion in human cells

David Liu and Richard Merkin of the Broad Institute of MIT and Harvard, along with Columbia professor Sam Sternberg, have developed a new, targeted means of inserting entire genes into human DNA (Science 2025, DOI: 10.1126/science.adt5199).

The new tool is called evoCAST, so named because it was created using a Liu-developed evolution technique on CRISPR-associated transposases (CASTs). CRISPR enables the targeting of precise sequences of DNA via a guide RNA—famously a core component of the CRISPR-Cas systems of gene editors. Transposases are enzymes that that cut and paste DNA within a genome; they’re the active component of some transposons, also known as jumping genes, which are so prolific that they make up roughly 50% of the human genome.

Shondra Pruett-Miller, a genome engineer at St. Jude’s Children’s Research Hospital who was not involved in the research, says that evoCAST is “a huge step forward” and that scientists have been working to create a CAST that can efficiently engineer human cells for about a decade.

CAST systems aren’t found in humans, and when used in human cell cultures they produce a successful DNA insertion only 0.1% of the time—far too low for gene therapy purposes. But after conducting hundreds of rounds of directed evolution, the team crafted a CAST that could make a successful insertion 10–20% of the time, which Pruett-Miller says could be efficient enough to be therapeutically useful for some diseases.

Other technologies for whole gene insertion exist, including other CAST- and transposase-based technologies. But many are limited in that they can inadvertently create mutations at the target site or can’t target specific insertion regions at all, according to Pruett-Miller.

The efficiency of evoCAST gives it a competitive advantage, though Pruett-Miller acknowledges there’s room for improvement: “I think this technology still needs a bit of refinement in order to make it to the clinic. The delivery is going to be a major issue.” The evoCAST system relies on a complex of proteins and a guide RNA, all of which can be hard to package and to deliver to specific cells—a common problem in many gene therapies.

The gene-editing field is still a developing space. The news of evoCAST’s development comes as Prime Medicine—cofounded by Liu to commercialize a different technology, Prime Editing—announced the successful demonstration of one of its therapies, PM359, in a patient with chronic granulomatous disease. But the same day, Prime changed CEOs, laid off a quarter of its staff, and stopped work on PM359 to focus on other areas of research.