Programmable large-DNA writing in the human genome.

Programmable genome editing has advanced from nucleases through base and prime editors, yielding an increasingly precise and versatile editing toolkit that is nonetheless best suited to small, localized changes. Efficient, site-specific writing of kilobase-scale DNA, including whole genes, regulatory elements, and multi-component cassettes, remains a central unmet goal, with implications for dissecting regulatory logic, building physiologically accurate disease models, and engineering mutation- and gene-agnostic therapies. Here, we review recent advances in programmable large-DNA writing in the human genome across four mechanistic categories: recombinase-, prime editing-, transposase-, and retrotransposon-mediated technologies. Across these platforms, high catalytic activity and precise control over the integration site have tended to trade off against one another, a tension further compounded by declining efficiency from immortalized lines to primary cells to in vivo models. We also discuss the key limitations facing efficient ex vivo and in vivo large-DNA writing applications, and consider how current and future advances in biomolecule evolution and delivery systems can help address these challenges.