In a recent study published in the journal Nature Biotechnology, researchers engineered prime editor (PE)-engineered virus-like particles (eVLPs) delivering PE proteins, PE guide ribonucleic acids (pegRNAs), and nicking single guide ribonucleic acids (ngRNAs) as ribonucleoprotein (RNP) complexes.
Study: Engineered virus-like particles for transient delivery of prime editor ribonucleoprotein complexes in vivo. Image Credit: Andrii Yalanskyi / Shutterstock
Background: The Promise of Prime Editing
Prime editing is a promising technology for changing genomic deoxyribonucleic acid (DNA) that has the potential to be used to cure genetic diseases in individuals. Prime editors are proteins that can replace a specific deoxyribonucleic acid sequence with another. PE systems necessitate three distinct nucleic acid hybridizations and are not dependent on double-strand deoxyribonucleic acid breaks or donor deoxyribonucleic acid templates.
Researchers must devise efficient and safe techniques to deliver prime editors in tissues in the in vivo settings to fulfill PE’s objective. While viral delivery techniques such as adenoviruses and adeno-associated viruses (AAVs) can transport PE in vivo, non-viral delivery techniques like lipid nanoparticles can sidestep these concerns by packaging PEs as temporarily expressing messenger ribonucleic acids.
Developing the PE-eVLP System
In the present study, researchers developed a prime editor-engineered VLP system to deliver prime editors, including ngRNAs and pegRNAs as ribonucleoproteins.
The team evaluated the PE-eVLP system in HEK293T cells, Neuro-2a cells, and Gesicle Producer 293T cells for cell culture tests. The researchers created v3 and v3b PE-eVLPs with 65- to 170-fold greater editing effectiveness in human cells than a previously reported base editor eVLP design. They used v1.2 prime editor-eVLPs with engineered PE guide ribonucleic acids (epegRNAs) and replaced the PE protein with PEmax2.
Optimizing for Efficiency: The v1.2 and v2.3 PE-eVLPs
The team used v1.2 PE-eVLPs with engineered pegRNAs to replace the prime editor protein with PEmax2, an enhanced PE that includes SpCas9 amino acid substitutions, an optimized linkage molecule between the RT domain, Cas9 nickase, nuclear localization signal (NLS) optimization, and codon optimization. They sought to identify mechanistic bottlenecks in v1 PE-eVLPs, solving the problem by relocating the nuclear export signals (NES) within the Gag protein and inserting three nuclear export signals before the site of protease cleavage of every Gag protein subdomain with two additional regions that could tolerate large insertions into the MMLV Gag-Pol.
The researchers observed that cellular mismatch repair (MMR) pathways can reduce PE efficiency and that avoiding or inhibiting MMR enhances PE efficiencies. To investigate this potential for eVLP-delivered primary editing systems, they used the v2.3 prime editor-eVLP system to insert additional close alterations at the HEK3 and Dnmt1 loci.
The researchers investigated whether the ngRNA could be packaged in the same or a different particle from the epegRNA to find the best all-in-one particle v3 PE3-eVLP system. They additionally examined whether these changes improved eVLP-mediated BE delivery. The transient introduction of PE via eVLPs reduced the capacity of v3 and v3b PE-eVLPs to facilitate in vivo prime editing in the mouse central nervous system.
Results: Advancements in PE-eVLP Technology
The research aims to create third-generation PE-eVLPs with clinically relevant levels of primary editing in the retina, protein expression restoration, and partial visual function rescue. The researchers found and designed prime editors and engineered VLP architectures, resulting in a PE efficiency boost of 79-fold in murine neuro-2A (N2A)-type cells and an improvement of 170-fold in human HEK293T cells compared to v1 PE-eVLPs. One subretinal v3 PE-eVLP injection corrected a 4.0-bp Mfrp deletion in the rd6 murine retinal degeneration model (mean efficiency of 15%) and an Rpe65 mutation to partially release visual functions in the rd12 model (mean efficiency of 7.2%).
Key Innovations in PE-eVLP Design
The nuclear export signals promote Gag-cargo protein localization in the cytoplasm, and the p= four additional MMLV protease cleavage regions in the Gag protein enhance incomplete cleavage possibility, leading to the retaining of some Gag and nuclear export signals by PE cargo. In the human embryonic kidney 293T cells, inserting three nuclear export signals between the CA and p12 domains of Gag (nuclear export signal position number 5) resulted in the highest PE efficiency, generating v2.2 prime editor-eVLP.
The researchers used the v2.3 prime editor-eVLP technique to insert extra adjacent replacements at the human embryonic kidney (HEK3) and Dnmt1 loci, which enhanced primary editing efficiency in both cases. Inadequate epegRNA packing hampered PE-eVLP efficiency, whereas epegRNA supplementation increased PE-engineered VLP editing efficiencies by more than 8.0-fold. With v3 PE3-eVLPs, the researchers achieved 2.3% bulk cortex editing and 36% editing among GFP+ nuclei.
Conclusion: Future Directions for PE-eVLPs
Overall, the study findings showed that optimized PE-eVLPs enable transitory in vivo administration of prime editor ribonucleoproteins, improving safety and inhibiting oncogenic transgene integration. In both culture and in vivo, these virus-like particles transport PE RNPs into mammalian cells. Recent advancements in primary editing systems, including epegRNAs, PEmax design, and MMR avoidance, have resulted in better results. In vivo, the improved v3 and v3b PE-eVLPs systems corrected pathogenic deletions in the mouse retina and achieved editing levels equivalent to triple-vector AAV-PE systems in genetic blindness models. Next-generation PEs and enhanced eVLP systems will need further technical work.