Site-specific DNA insertion into the human genome with engineered recombinases

Large serine recombinases (LSRs) offer a promising approach for site-specific integration of multi-kilobase DNA sequences into genomes. We present a comprehensive strategy for engineering LSRs to enhance both efficiency and specificity of genome integration without pre-installed landing pads. Using Dn29 as a model, we combined directed evolution, structural analysis, and machine learning to identify and optimize mutations affecting efficiency and specificity. We further improved performance through dCas9 fusions for programmable target and donor recruitment and donor DNA optimization. The engineered Dn29 variants achieved up to 55% integration efficiency and 97% genome-wide specificity at an endogenous human locus. These recombinases effectively integrated large DNA cargoes (up to 12 kb) for stable expression in challenging cell types, including non-dividing cells, human embryonic stem cells, and primary T cells. Our approach provides a blueprint for rational engineering of recombinases, enabling precise genomic engineering and expanding the potential for both research and therapeutic applications requiring large DNA insertions.