Programmable bridge recombinases enable megabase‑scale edits in the human genome

Introduction

A technology that can rearrange nearly an entire chromosome arm — cleanly, predictably, and without the blunt trauma of double‑strand breaks — sounds like science fiction. Yet a set of papers published between 2024 and 2025 describes exactly that: a family of enzymes guided by a bispecific RNA scaffold, now adapted for human cells to perform programmed megabase‑scale genome rearrangements (e.g., insertions, excisions, inversions) (Perry et al., Science 2025).

What is a bridge recombinase?

Bridge recombinases are site‑specific enzymes that use a structured noncoding "bridge RNA" with two independent guide loops. One loop base‑pairs to a genomic target; the other base‑pairs to a donor sequence. The RNA brings the two DNA loci into the enzyme's active site so the recombinase can perform strand exchange and religation — producing scarless rearrangements without leaving long double‑strand breaks behind (Durrant et al., Nature 2024; Hiraizumi et al., Nature 2024).

Key findings at a glance

• Human‑adapted system: Researchers engineered a human‑active variant called IS622 and optimized its bridge RNAs and protein sequence to function in human cells (Perry et al., Science 2025).
• Scale and efficiency: The team demonstrated intra‑chromosomal inversions and excisions mobilizing sequences up to ~0.93 megabases, with insertion efficiencies approaching ~20% in their assays and genome‑wide specificity measures up to ~82% (Perry et al., Science 2025).
• Foundational biology and mechanism: The original discovery showed bispecific guide RNAs in the IS110 family can be reprogrammed to direct recombination between chosen DNA partners with high efficiency in bacteria (>60% in some assays) and high specificity (>90%) (Durrant et al., Nature 2024).
• Structural validation: Cryo‑EM structures reveal a synaptic complex where two recombinase dimers and the bridge RNA coordinate cleavage and strand exchange in a way that favors religation and minimizes free double‑strand breaks (Hiraizumi et al., Nature 2024).

How the mechanism changes the genome‑editing playbook

Most popular genome editors (nucleases or base editors) target one site and rely on cellular repair pathways, sometimes generating undesired outcomes. Bridge recombinases are intrinsically bispecific: they scaffold two sequences and catalyze strand exchange, enabling scarless insertions or rearrangements without the same dependence on host homology‑directed repair (Durrant et al., Nature 2024). Structural work explains how the protein–RNA complex orchestrates strand cleavage and religation to produce clean junctions (Hiraizumi et al., Nature 2024).

Real examples and potential applications

• Therapeutic excision: Perry and colleagues demonstrated proof‑of‑concept excision of regulatory regions and pathogenic repeat expansions relevant to disease, illustrating a path toward therapies that remove harmful DNA sequences rather than trying to patch them (Perry et al., Science 2025).
• Synthetic chromosome design: Scarless insertions and large inversions are valuable for building synthetic chromosome segments or reprogramming regulatory architecture in stem cells (Durrant et al., Nature 2024).
• Functional genomics: Researchers can create large, defined genomic deletions or inversions to study locus‑scale gene regulation or 3D chromatin effects with much cleaner outcomes than iterative nuclease strategies.

Why this matters

• Precision at scale: The ability to move or remove nearly megabase‑scale regions with high fidelity opens experimental and therapeutic possibilities that were previously slow, imprecise, or impractical.
• Reduced collateral damage: By avoiding persistent double‑strand breaks and stitching DNA back together, bridge recombinases may lower genomic instability and unintended rearrangements compared with some nuclease approaches (Hiraizumi et al., Nature 2024).
• New design paradigm: Bispecific RNA guides expand the concept of programmable nucleic acid guides from single‑site targeting (CRISPR) to bridging two loci, enabling operations (excision, insertion, inversion) in one coordinated step (Durrant et al., Nature 2024).

Challenges, unknowns and ethical questions

• delivery and efficiency: Reported human‑cell efficiencies (up to ~20%) and specificity (~82%) are promising but need improvement for many therapeutic applications; delivering large recombinase complexes and donor DNA to patient tissues remains a hurdle (Perry et al., Science 2025).
• off‑target and long‑term safety: Even with structural‑guided engineering to reduce off‑targets, megabase rearrangements have system‑level consequences—chromatin context, regulatory networks, and 3D genome folding could produce unanticipated effects.
• governance and ethics: What governance frameworks are required when we can rewrite chromosome‑scale architecture? Who decides which genomic rearrangements are therapeutic vs. experimental enhancement? These are not technical issues alone.

Questions to ponder

• If we can cleanly excise a pathogenic repeat expansion, how should we prioritize diseases to target first?
• What safeguards should be standard before attempting megabase edits in germline cells or embryos?

Limitations and the path forward

The authors are transparent about limits: efficiency and specificity need improvement in diverse, clinically relevant cell types; delivery remains nontrivial; and pushing beyond ~1 Mb will require further enzyme and RNA engineering (Perry et al., Science 2025). Structural maps from cryo‑EM give concrete mutation sites and RNA designs that can be iteratively optimized (Hiraizumi et al., Nature 2024).

Conclusion and call to action

Bridge recombinases represent a conceptual and technical leap: they turn a noncoding RNA into a programmable scaffold that brings two DNA sequences together and writes the genome at a scale previously cumbersome to change. This line of work — discovered and mechanistically explained in 2024 and adapted for human cells in 2025 — is rightly called revolutionary because it reframes what ‘‘targeted’’ editing can mean (Durrant et al., Nature 2024; Hiraizumi et al., Nature 2024; Perry et al., Science 2025).

Concrete next steps for the field include improving delivery, benchmarking safety across cell types, and creating international standards for responsible research. For researchers and funders, the question is not just whether we can do megabase edits, but how we should do them in ways that maximize benefit and minimize harm.

Original source: https://www.science.org/doi/10.1126/science.adz0276

References

• Perry NT et al., "Megabase‑scale human genome rearrangement with programmable bridge recombinases," Science (2025). https://www.science.org/doi/10.1126/science.adz0276
• Durrant MG et al., "Bridge RNAs direct modular and programmable recombination of target and donor DNA," Nature (2024). https://pubmed.ncbi.nlm.nih.gov/38328150/
• Hiraizumi M et al., "Structural mechanism of bridge RNA‑guided recombination," Nature (2024). https://doi.org/10.1038/s41586-024-07570-2