Published by Emerging Technologies Laboratory · via ETL Newswire
Science· 

Columbia Preprint Reports Efficient Base Editing of Human Embryos Without Chromosomal Damage

A bioRxiv preprint from Dieter Egli's lab shows adenine base editors can correct disease-linked variants in human embryos without triggering the large deletions that plagued earlier CRISPR attempts - but the work is not peer reviewed and remains years from any clinical use.

By Dr. Maya Iyer, Staff Reporter · Science Desk

A team led by developmental cell biologist Dieter Egli at Columbia University posted a preprint to bioRxiv on June 1 describing what they say is efficient base editing of human embryos at two clinically relevant genomic targets, PCSK9 and HBG, without producing the chromosomal abnormalities that have undermined previous gene-editing attempts in embryonic cells.

The significance of the chromosomal finding requires some context. Standard CRISPR-Cas9 editing works by introducing deliberate double-strand breaks in DNA, then relying on the cell's own repair machinery to fix the cut. In adult somatic cells that approach works reasonably well. In early human embryos, it does not. As the preprint itself notes, those double-strand breaks result in frequent aneuploidy and large deletions, reflecting a repair deficiency specific to early embryonic development and one that had effectively ruled out Cas9-based editing as a clinical path.

Base editing sidesteps that problem. According to the bioRxiv preprint, the technique converts individual DNA letters - in this case adenine to guanine - without creating double-strand breaks. The Egli team used adenine base editors delivered as ribonucleoprotein complexes and found that editing was efficient at both targets, that small insertions or deletions were rare, and that off-target activity depended on the guide RNA used. Crucially, unlike Cas9-induced cuts, the base edits did not result in either chromosomal abnormalities or large deletions. Edited embryos developed normally to the blastocyst stage and yielded stem-cell lines carrying the intended edits.

The two targets are not arbitrary. PCSK9 is linked to cholesterol regulation; variants that reduce PCSK9 function are associated with markedly lower cardiovascular risk. HBG1 and HBG2 encode fetal hemoglobin subunits relevant to sickle cell disease and beta-thalassemia. Introducing gain-of-function edits in those genes in embryos would, in theory, create heritable protection against those conditions - which is precisely why the work has attracted both interest and alarm.

Reporting in Nature News described the preprint as "a conceptual shift" in the field, citing Yale fertility specialist Emre Seli. But the same coverage notes that the technique is far from ready for the clinic, and critics have pointed to the timing relative to an active moratorium. The American Society of Gene and Cell Therapy, along with the International Society of Cell and Gene Therapy and the Alliance for Regenerative Medicine, called in May 2025 for a 10-year ban on heritable human genome editing. As C&EN reported, ASGCT CEO David Barrett said the Columbia work "flies in the face" of that moratorium.

The preprint's own authors hedge accordingly. Their closing line, as reviewed in C&EN, reads: "Though this may be a step towards heritable editing, translation to a clinical context remains premature."

Several methodological caveats deserve attention before this work is interpreted too broadly. The preprint has not yet been peer reviewed. The authors note that off-target editing rates depend heavily on guide RNA design, which means the safety profile seen here is not a universal property of base editing but a function of the specific constructs used. The study also does not report embryo numbers in the summary abstract, a detail that matters considerably when evaluating efficiency claims in a system where obtaining human embryos for research is ethically constrained and logistically limited.

The preprint pipeline is doing exactly what it should here: getting results into scientific circulation quickly so the field can scrutinize them. But the gap between a bioRxiv posting and peer-reviewed, replicated data is real, and in a domain where the regulatory and ethical stakes are this high, that gap is not a formality.

Sources cited:
- bioRxiv preprint - Jerabek, Egli et al. (https://www.biorxiv.org/content/10.64898/2026.05.30.728989v1)
- Chemical & Engineering News (https://cen.acs.org/biological-chemistry/gene-editing/base-editing-human-embryos/104/web/2026/06)
- Nature News (https://www.nature.com/articles/d41586-026-01827-8)
- CRISPR Medicine News (https://crisprmedicinenews.com/news/base-editing-rewrites-embryo-genomes-precisely/)

Reporting by Dr. Maya Iyer, Staff Reporter, for the Science desk · ETL Newswire staff
Read more at the source

This release was originally distributed via ETL Newswire. Visit bioRxiv preprint - Jerabek, Egli et al. for the full story, related releases, and contact information.

Visit bioRxiv preprint - Jerabek, Egli et al. →