Retrotransposons: A New Frontier in In Vivo Genome Engineering

Genome engineering has transformed cell therapy, but clinical implementation remains largely confined to ex vivo workflows followed by adoptive cell transfer. These approaches require costly manufacturing, often rely on viral vectors, and carry risks associated with unwanted genomic alterations. Recent advances in lipid nanoparticles (LNPs) enable genome-editing components to be delivered directly to specific cell populations in vivo, eliminating the need for ex vivo manufacturing. Several RNA-based genome-editing approaches have entered clinical trials. Nevertheless, current in vivo approaches remain largely limited to gene disruption or small sequence modifications. Conventional integration of large payloads generally requires DNA donors, which can be immunogenic, have limited nuclear access, and pose substantial LNP formulation challenges. When expression of therapeutic proteins such as chimeric antigen receptors (CARs) is required, LNP-based approaches predominantly deliver mRNA, leading to transient expression, whereas many applications require durable expression through stable genomic integration. Retrotransposons offer a solution because they have naturally evolved to reverse-transcribe RNA templates directly into chromosomal DNA. Engineered retrotransposon-derived platforms can integrate multi-kilobase payloads from RNA without an exogenous DNA donor. These platforms combine durable expression with RNA-only delivery, but clinical translation requires safe permanent integration, copy-number control, efficient delivery of all required components, and sufficient activity in disease-relevant primary cells. In this review, we describe the classes of non-LTR retrotransposons; examine the design principles underlying their adaptation as genome-editing tools; and compare emerging platforms derived from R2, LINE-1, Vingi, and RTE elements. These systems support integration strategies ranging from naturally evolved insertion profiles, such as site-specific insertion into ribosomal DNA (rDNA) or dispersed genome-wide integration, to fully programmable insertion. Several retrotransposon-based platforms have demonstrated multi-kilobase payload insertion, functional engineering of primary human T cells, and in vivo activity in mouse or non-human primate models, demonstrating their clinical promise. We also discuss key translational challenges. With continued optimization, retrotransposon-based platforms have the potential to bridge the gap between transient RNA delivery and durable gene writing, establishing a versatile foundation for the next generation of in vivo genetic medicines.