Bridge RNA Reveals the “One-Way Street” of DNA Insertion: IS621 Structures Point to Programmable Large-Fragment Editing

For gene editing to insert an entire gene or a large regulatory sequence at once, the key is not only to “cut precisely,” but also to deliver a large DNA fragment to a specified location. A study published in *Nature* further dissects the bacterial IS621 bridge recombination system, explaining why it is naturally biased toward insertion rather than excision and offering RNA design clues for altering the direction of the reaction.

IS621 belongs to the IS110 family of mobile genetic elements. It must first be excised from the bacterial chromosome to form circular DNA and then insert into a new location. The bridge RNA responsible for recognizing the two sets of DNA contains regions that can pair separately with the target and donor sequences. The problem is that the bridge RNA’s strong promoter becomes fully formed only after the element circularizes, seemingly creating a “which comes first—the RNA or completed excision?” loop.

In a commonly used *Escherichia coli* strain containing three copies of IS621, the research team found that linear elements can still produce bridge RNA and recombinase through low levels of readthrough transcription, sufficient to initiate rare excision events. A research briefing released by the University of Tokyo also focused on this mechanism: a small amount of RNA first allows the element to leave the chromosome; the promoter reconstructed after circularization then increases RNA and recombinase expression, driving subsequent insertion.

Biochemical experiments showed that the natural system is highly asymmetric in its directionality. Under one set of in vitro conditions, excision efficiency was approximately 0.03%, while insertion reached 63%, a difference of about 2,000-fold. Cryo-electron microscopy structures provided a physical explanation: during insertion, the donor and target DNA bend into a U shape; during excision, the two DNA segments are relatively straight and cross in an X shape. Both reactions can complete the initial cleavage, but excision is more readily impeded during the subsequent strand-exchange step.

Short “handshake guide sequences” in the bridge RNA act like directional switches, stabilizing the DNA state before or after exchange through base pairing. After rewriting these sequences and extending some guide regions, the researchers increased excision efficiency by approximately 13,000-fold relative to the natural bridge RNA in bacterial plasmid experiments. Simply increasing RNA expression provided no further benefit, indicating that sequence design is more important than producing large amounts of RNA.

Background

The appeal of bridge RNA recombination systems lies in the ability of a single RNA molecule to specify both donor and target DNA, theoretically making them suitable for large-fragment insertion without relying entirely on the cell to repair breaks on its own. However, this study mainly consists of experiments in bacteria, in vitro reactions using purified molecules, and structural analyses. The large-fold improvement also came from a specific plasmid model and cannot be directly translated into efficiency in the human genome. To become a practical tool, delivery, sequence selectivity, unintended rearrangements, and long-term safety in mammalian cells will still need to be validated.