Researchers led by Daniel E. Bauer of Boston Children’s Hospital and Harvard Medical School, with first author Sébastien Levesque, describe “prime assembly,” a genome-engineering method that uses CRISPR-targeted dual flap synthesis to assemble and integrate medium-to-large DNA fragments at programmed sites in human cells. The peer-reviewed paper, “Targeted genomic integration and rearrangement using prime assembly,” was published in Nature (opens in new tab) on Sept. 16, 2026; a Harvard Medical School summary of the work was republished Oct. 1.

Unlike standard CRISPR-Cas9 knock-ins that rely on double-strand breaks and homology-directed repair, prime assembly builds on prime editing’s single-strand nicks. The authors report activity in dividing and non-dividing cells, including primary human T cells and CD34+ hematopoietic stem and progenitor cells, and demonstrate exon recoding, kilobase-scale transgene integration and megabase-scale rearrangements in experimental systems.

Co-senior author Bauer said the approach controls where DNA replacement starts and ends and is “much less likely to cause off-target effects compared to other gene-editing methods,” according to an HMS-sourced summary republished by Technology.org. The team is working on delivering components to disease-relevant cells in vivo, such as hematopoietic stem cells for blood disorders. Levesque and Bauer have filed a patent application covering the technology, according to the paper’s competing-interests statement.

The results are preclinical laboratory findings. Efficiencies vary by locus and donor design; imprecise alleles still occur, though DNA-PK and Polθ inhibitors improved purity in reported experiments. Clinical use would require delivery advances, further safety studies and regulatory review — none of which are claimed as complete in the paper.