Machine learning prediction of prime editing efficiency across diverse chromatin contextsNicolas Mathis, Ahmed Allam, András Tálas et al.|Nature Biotechnology|2024 The success of prime editing depends on the prime editing guide RNA (pegRNA) design and target locus. Here, we developed machine learning models that reliably predict prime editing efficiency. PRIDICT2.0 assesses the performance of pegRNAs for all edit types up to 15 bp in length in mismatch repair-deficient and mismatch repair-proficient cell lines and in vivo in primary cells. With ePRIDICT, we further developed a model that quantifies how local chromatin environments impact prime editing rates. A machine learning model for prime editing efficiency prediction takes into account chromatin context.
Treatment of a metabolic liver disease in mice with a transient prime editing approachAbstract Prime editing is a versatile genome editing technology that circumvents the need for DNA double-strand break formation and homology-directed repair, making it particularly suitable for in vivo correction of pathogenic mutations. Here we developed liver-specific prime editing approaches with temporally restricted prime editor (PE) expression. We first established a dual-delivery approach where the prime editor guide RNA is continuously expressed from adeno-associated viral vectors and only the PE is transiently delivered as nucleoside-modified mRNA encapsulated in lipid nanoparticles (LNP). This strategy achieved 26.2% editing with PEmax and 47.4% editing with PE7 at the Dnmt1 locus using a single 2 mg kg −1 dose of mRNA–LNP. When targeting the pathogenic Pah enu2 mutation in a phenylketonuria mouse model, gene correction rates reached 4.3% with PEmax and 20.7% with PE7 after three doses of 2 mg kg −1 mRNA–LNP, effectively reducing blood l -phenylalanine levels from over 1,500 µmol l −1 to below the therapeutic threshold of 360 µmol l −1 . Encouraged by the high efficiency of PE7, we next explored a simplified approach where PE7 mRNA was co-delivered with synthetic prime editor guide RNAs encapsulated in LNP. This strategy yielded 35.9% editing after two doses of RNA–LNP at the Dnmt1 locus and 8.0% editing after three doses of RNA–LNP at the Pah enu2 locus, again reducing l -phenylalanine levels below 360 µmol l −1 . These findings highlight the therapeutic potential of mRNA–LNP-based prime editing for treating phenylketonuria and other genetic liver diseases, offering a scalable and efficient platform for future clinical translation.
Publisher Correction: Machine learning prediction of prime editing efficiency across diverse chromatin contextsNicolas Mathis, Ahmed Allam, András Tálas et al.|Nature Biotechnology|2024 Spatial profiling of gene editing by in situ sequencing in mice and macaquesStrategies for Interdisciplinary Human Gene Editing Research: Insights from a Swiss ProjectCRISPR gene editing is a cutting-edge technology that has advanced tremendously in recent years. The first clinical CRISPR applications have been approved, and more gene editing therapies are to be expected in human medicine. Consequently, continuous basic research is needed to assess possibilities and prime future clinical applications. Because this technology not only offers new possibilities for treating diseases but also raises important ethical and societal questions, collaboration between human, life, biomedical, and medical sciences is needed. In this article, we discuss the practical challenges of such interdisciplinary projects and present strategies for addressing them based on our experience of conducting an interdisciplinary project on CRISPR. This work aims to help and encourage interdisciplinary collaborations and discussions on modern scientific endeavors that, such as gene editing, tend to blur the lines between traditional disciplines. The strategies suggested include realistic expectations, shared goals, space setting, and expert and lay dialogue.