Il Prime Editing Gemello Potenziato Riduce i Ripetizioni Geniche Difettose nelle Cellule Modello di Atassia

Spinocerebellar ataxia type 3 (SCA3) is caused by an abnormally expanded CAG repeat in the ATXN3 gene, producing a toxic polyglutamine (polyQ)-containing protein. Correcting such large genomic alterations exceeds the capacity of standard CRISPR base editors or single prime editors, making twin prime editing — which deploys two coordinated prime editor complexes to generate complementary DNA flaps — an attractive but still-immature strategy.

This study set out to improve twin prime editing on two fronts: protein engineering and guide RNA design. The team fused the La protein domain (residues 1–194), known to shield RNA 3′-ends from exonucleolytic degradation, to the SpCas9(H840A)-reverse transcriptase prime editor, creating La-twin-PE. They also systematically compared three pegRNA overlap architectures — whole overlap (fully complementary), homology arm, and partial overlap — across three founding genomic loci (HEK3, FANCF, AAVS1) in HEK293FT cells, then scaled testing to 13 loci.

Key results showed that La-twin-PE combined with whole-overlap or partial-overlap pegRNA designs consistently outperformed the conventional twin-PE, yielding a statistically significant 1.75 ± 0.21-fold increase in precise editing across 13 loci. Crucially, unintended indel rates remained comparably low, preserving fidelity. Homology-arm pegRNA designs, by contrast, did not benefit from La fusion and actually generated higher rates of indel-only outcomes, especially with wild-type SpCas9 modules. Unexpectedly, combining the La-twin-PE module with epegRNAs (which carry a 3′ pseudoknot stabilization element) suppressed editing efficiency, suggesting the two RNA-protection mechanisms are mechanistically incompatible.

Beyond small-sequence insertion, La-twin-PE enabled efficient Bxb1 recombinase-mediated knockin of a ~2.8 kb GFP transgene at target loci following attB/attP site installation, demonstrating scalability to therapeutically relevant cargo sizes. The most disease-relevant result was in HEK293FT-ATXN3-Q84 cells, an engineered human cell model of SCA3: La-twin-PE successfully eliminated the expanded polyQ-encoding repeat tract in ATXN3, providing proof-of-concept for therapeutic genome correction of repeat-expansion disorders.

While the findings are promising, the study is currently limited to cultured human cell lines; efficacy and safety in post-mitotic neurons or in vivo animal models remain to be established. Delivery of the large La-twin-PE construct to the brain presents an additional translational hurdle. Nonetheless, the optimized platform offers a more efficient and streamlined route to large-scale, precise genome editing relevant to SCA3, Huntington disease, hemophilia, and other complex genetic conditions.