Optimized lipid nanoparticle RNA design achieves 49% prime editing in mouse liver
Background: Coordinating the delivery and activity of three RNAs
A study published in Nature Nanotechnology achieved an average of 49% prime editing at the Pcsk9 target in mouse liver by optimizing the RNAs packaged in lipid nanoparticles (LNPs) in stages. In a separate experiment in phenylketonuria (PKU) mice, the researchers corrected the disease-causing variant and observed a reduction in serum phenylalanine. These findings came from preclinical experiments with different targets and doses.
Prime editing introduces base substitutions or short insertions or deletions at specified DNA sites. The delivery system in this study comprised messenger RNA (mRNA) encoding the editing enzyme, prime editing guide RNA (pegRNA) carrying the correction instructions, and a guide RNA (ngRNA) that induces an additional single-strand DNA nick. This creates a timing constraint: the guide RNAs must remain sufficiently available until the enzyme is produced from the mRNA.
An earlier PE7-LNP study cited in the paper reported up to 23% editing in bulk mouse liver after two doses of 4 mg/kg. The present study focused on increasing the efficiency of a single dose by testing RNA components in sequence. Results from different studies need to be considered alongside their dosing conditions.
Key findings: Pcsk9 editing and correction of a PKU variant
The researchers encapsulated each RNA separately in OF-02-based LNPs, then mixed them immediately before administration. Their experimental optimization involved sequentially testing the editing enzyme, the 3′-terminal motif of an engineered pegRNA (epegRNA) designed to reduce degradation, and the RNA mass ratio, while also improving RNA purification quality.
For the Pcsk9 target in healthy mice, they used PE6c and the eSBRMV1-A motif. Seven days after a single intravenous dose of 2 mg/kg total RNA, the intended TTAC insertion was detected at an average frequency of 49% in bulk liver DNA. This was 63 times the result of the initial approach at the same dose. Bulk liver contains a mixture of cell types, so this figure should not be interpreted as the editing rate in isolated hepatocytes. The choice of enzyme and motif had a substantial effect, whereas differences among the RNA ratios tested were comparatively limited.
Correction of the disease-causing variant was assessed separately in 7–9-week-old PKU mice homozygous for the humanized PAH R408W variant. The PAH variant causes phenylalanine to accumulate to neurotoxic levels. Four mice per group received a single dose of 4 mg/kg total RNA. At day 7, correction rates in bulk liver genomic DNA were 15% and 12% for the s2 and s3 combinations, respectively. The corresponding correction rates in complementary DNA (cDNA) made from transcripts were 23% and 19%. The final optimization step, s3, did not produce the highest efficiency at every target.
With the s2 combination, serum phenylalanine fell by 90% within three days of dosing. By day 7, every PE-LNP treatment group except the initial s0.2 combination had levels below 360 µmol/L, the therapeutic intervention threshold cited in the paper. The researchers interpreted the result as reaching a level at which therapeutic benefit might be expected, but metabolic markers in this experiment were followed only through day 7. The study did not demonstrate a cure in humans or long-term neurological benefit.
Implications and outlook: Repeat dosing and the scope of safety assessment
In the Pcsk9 experiment, two doses of 1 mg/kg given seven days apart without steroid pretreatment increased editing from 28% after one dose to 43% after two. In a separate 2 mg/kg experiment, a mild increase in serum alanine aminotransferase (ALT) had returned to a level similar to the control group three days after dosing.
Off-target editing comparisons should be interpreted within the set of candidate sites examined. In mice, 14 Pcsk9-related candidate sites were assessed eight weeks after dosing. Off-target editing at OT13 was observed with both LNP and adeno-associated virus (AAV) delivery, at a lower frequency in the LNP group. These findings alone cannot establish the absence of off-target editing across the whole genome or long-term safety in humans.
The results provide evidence that target-specific RNA design and purification quality can improve nonviral prime editing. Application in humans will require evaluation of durability, appropriate dosing, immune responses, and a broader range of safety outcomes. Detailed experimental conditions and results are available in the full paper.
Prime editing is a versatile clinical genome editing method that enables precise substitutions, small insertions and deletions at specified locations in the genomes of living systems including human cells. Although non-viral lipid nanoparticle (LNP) delivery of RNA in vivo has become a preferred method for gene editing in animals and patients, its application to complex, three-component prime editing systems has yielded low editing efficiencies. Here we developed a systematic prime editing LNP (PE-LNP) optimization platform that addresses key bottlenecks in cargo design that limit editing efficiency. This generalizable workflow yielded PE-LNPs that can achieve 49% average in vivo prime editing in the bulk mouse liver with a single dose of 2 mg kg
For researchers developing prime editing candidates for inherited liver diseases, the study offers a concrete starting point for target-specific optimization. The approach is to select an editing enzyme and epegRNA motif in cultured cells carrying the disease variant, then test the mixing ratio and purification quality of separately manufactured RNA-LNPs, and assess correction rates alongside metabolic markers in an animal model.
The finding that suitable enzymes and editing performance differed between Pcsk9 and PAH shows why each target’s combination needs to be validated anew. In follow-up development, this workflow can be used to adjust RNA purity and delivery conditions while checking whether DNA correction translates into improved metabolic function. The short-term fall in phenylalanine in PKU mice supports further development, but does not establish long-term treatment efficacy or prevention of neurological damage in patients.