Effective genome editing with an enhanced ISDra2 TnpB system and deep learning-predicted ωRNAs.

Transposon (IS200/IS605)-encoded TnpB proteins are predecessors of class 2 type V CRISPR effectors and have emerged as one of the most compact genome editors identified thus far. Here, we optimized the design of Deinococcus radiodurans (ISDra2) TnpB for application in mammalian cells (TnpBmax), leading to an average 4.4-fold improvement in editing. In addition, we developed variants mutated at position K76 that recognize alternative target-adjacent motifs (TAMs), expanding the targeting range of ISDra2 TnpB. We further generated an extensive dataset on TnpBmax editing efficiencies at 10,211 target sites. This enabled us to delineate rules for on-target and off-target editing and to devise a deep learning model, termed TnpB editing efficiency predictor (TEEP; https://www.tnpb.app ), capable of predicting ISDra2 TnpB guiding RNA (ωRNA) activity with high performance (r > 0.8). Employing TEEP, we achieved editing efficiencies up to 75.3% in the murine liver and 65.9% in the murine brain after adeno-associated virus (AAV) vector delivery of TnpBmax. Overall, the set of tools presented in this study facilitates the application of TnpB as an ultracompact programmable endonuclease in research and therapeutics.

Continuous directed evolution of a compact CjCas9 variant with broad PAM compatibility.

CRISPR-Cas9 genome engineering is a powerful technology for correcting genetic diseases. However, the targeting range of Cas9 proteins is limited by their requirement for a protospacer adjacent motif (PAM), and in vivo delivery is challenging due to their large size. Here, we use phage-assisted continuous directed evolution to broaden the PAM compatibility of Campylobacter jejuni Cas9 (CjCas9), the smallest Cas9 ortholog characterized to date. The identified variant, termed evoCjCas9, primarily recognizes N4AH and N5HA PAM sequences, which occur tenfold more frequently in the genome than the canonical N3VRYAC PAM site. Moreover, evoCjCas9 exhibits higher nuclease activity than wild-type CjCas9 on canonical PAMs, with editing rates comparable to commonly used PAM-relaxed SpCas9 variants. Combined with deaminases or reverse transcriptases, evoCjCas9 enables robust base and prime editing, with the small size of evoCjCas9 base editors allowing for tissue-specific installation of A-to-G or C-to-T transition mutations from single adeno-associated virus vector systems.

In vivo base editing of a pathogenic Eif2b5 variant improves vanishing white matter phenotypes in mice.

Vanishing white matter (VWM) is a fatal leukodystrophy caused by recessive mutations in subunits of the eukaryotic translation initiation factor 2B. Currently, there are no effective therapies for VWM. Here, we assessed the potential of adenine base editing to correct human pathogenic VWM variants in mouse models. Using adeno-associated viral vectors, we delivered intein-split adenine base editors into the cerebral ventricles of newborn VWM mice, resulting in 45.9% ± 5.9% correction of the Eif2b5R191H variant in the cortex. Treatment slightly increased mature astrocyte populations and partially recovered the integrated stress response (ISR) in female VWM animals. This led to notable improvements in bodyweight and grip strength in females; however, locomotor disabilities were not rescued. Further molecular analyses suggest that more precise editing (i.e., lower rates of bystander editing) as well as more efficient delivery of the base editors to deep brain regions and oligodendrocytes would have been required for a broader phenotypic rescue. Our study emphasizes the potential, but also identifies limitations, of current in vivo base-editing approaches for the treatment of VWM or other leukodystrophies.

Introducing gene deletions by mouse zygote electroporation of Cas12a/Cpf1.

CRISPR-associated (Cas) nucleases are established tools for engineering of animal genomes. These programmable RNA-guided nucleases have been introduced into zygotes using expression vectors, mRNA, or directly as ribonucleoprotein (RNP) complexes by different delivery methods. Whereas microinjection techniques are well established, more recently developed electroporation methods simplify RNP delivery but can provide less consistent efficiency. Previously, we have designed Cas12a-crRNA pairs to introduce large genomic deletions in the Ubn1, Ubn2, and Rbm12 genes in mouse embryonic stem cells (ESC). Here, we have optimized the conditions for electroporation of the same Cas12a RNP pairs into mouse zygotes. Using our protocol, large genomic deletions can be generated efficiently by electroporation of zygotes with or without an intact zona pellucida. Electroporation of as few as ten zygotes is sufficient to obtain a gene deletion in mice suggesting potential applicability of this method for species with limited availability of zygotes.

Gene editing of NCF1 loci is associated with homologous recombination and chromosomal rearrangements.

CRISPR-based genome editing of pseudogene-associated disorders, such as p47phox-deficient chronic granulomatous disease (p47 CGD), is challenged by chromosomal rearrangements due to presence of multiple targets. We report that interactions between highly homologous sequences that are localized on the same chromosome contribute substantially to post-editing chromosomal rearrangements. We successfully employed editing approaches at the NCF1 gene and its pseudogenes, NCF1B and NCF1C, in a human cell line model of p47 CGD and in patient-derived human hematopoietic stem and progenitor cells. Upon genetic engineering, a droplet digital PCR-based method identified cells with altered copy numbers, spanning megabases from the edited loci. We attributed the high aberration frequency to the interaction between repetitive sequences and their predisposition to recombination events. Our findings emphasize the need for careful evaluation of the target-specific genomic context, such as the presence of homologous regions, whose instability can constitute a risk factor for chromosomal rearrangements upon genome editing.

Machine learning prediction of prime editing efficiency across diverse chromatin contexts.

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.

Enhancing prime editor activity by directed protein evolution in yeast.

Prime editing is a highly versatile genome editing technology that enables the introduction of base substitutions, insertions, and deletions. However, compared to traditional Cas9 nucleases prime editors (PEs) are less active. In this study we use OrthoRep, a yeast-based platform for directed protein evolution, to enhance the editing efficiency of PEs. After several rounds of evolution with increased selection pressure, we identify multiple mutations that have a positive effect on PE activity in yeast cells and in biochemical assays. Combining the two most effective mutations - the A259D amino acid substitution in nCas9 and the K445T substitution in M-MLV RT - results in the variant PE_Y18. Delivery of PE_Y18, encoded on DNA, mRNA or as a ribonucleoprotein complex into mammalian cell lines increases editing rates up to 3.5-fold compared to PEmax. In addition, PE_Y18 supports higher prime editing rates when delivered in vivo into the liver or brain. Our study demonstrates proof-of-concept for the application of OrthoRep to optimize genome editing tools in eukaryotic cells.

Predicting prime editing efficiency and product purity by deep learning.

Prime editing is a versatile genome editing tool but requires experimental optimization of the prime editing guide RNA (pegRNA) to achieve high editing efficiency. Here we conducted a high-throughput screen to analyze prime editing outcomes of 92,423 pegRNAs on a highly diverse set of 13,349 human pathogenic mutations that include base substitutions, insertions and deletions. Based on this dataset, we identified sequence context features that influence prime editing and trained PRIDICT (prime editing guide prediction), an attention-based bidirectional recurrent neural network. PRIDICT reliably predicts editing rates for all small-sized genetic changes with a Spearman's R of 0.85 and 0.78 for intended and unintended edits, respectively. We validated PRIDICT on endogenous editing sites as well as an external dataset and showed that pegRNAs with high (>70) versus low (<70) PRIDICT scores showed substantially increased prime editing efficiencies in different cell types in vitro (12-fold) and in hepatocytes in vivo (tenfold), highlighting the value of PRIDICT for basic and for translational research applications.

Recursive Editing improves homology-directed repair through retargeting of undesired outcomes.

CRISPR-Cas induced homology-directed repair (HDR) enables the installation of a broad range of precise genomic modifications from an exogenous donor template. However, applications of HDR in human cells are often hampered by poor efficiency, stemming from a preference for error-prone end joining pathways that yield short insertions and deletions. Here, we describe Recursive Editing, an HDR improvement strategy that selectively retargets undesired indel outcomes to create additional opportunities to produce the desired HDR allele. We introduce a software tool, named REtarget, that enables the rational design of Recursive Editing experiments. Using REtarget-designed guide RNAs in single editing reactions, Recursive Editing can simultaneously boost HDR efficiencies and reduce undesired indels. We also harness REtarget to generate databases for particularly effective Recursive Editing sites across the genome, to endogenously tag proteins, and to target pathogenic mutations. Recursive Editing constitutes an easy-to-use approach without potentially deleterious cell manipulations and little added experimental burden.

In vivo adenine base editing of PCSK9 in macaques reduces LDL cholesterol levels.

Most known pathogenic point mutations in humans are C•G to T•A substitutions, which can be directly repaired by adenine base editors (ABEs). In this study, we investigated the efficacy and safety of ABEs in the livers of mice and cynomolgus macaques for the reduction of blood low-density lipoprotein (LDL) levels. Lipid nanoparticle-based delivery of mRNA encoding an ABE and a single-guide RNA targeting PCSK9, a negative regulator of LDL, induced up to 67% editing (on average, 61%) in mice and up to 34% editing (on average, 26%) in macaques. Plasma PCSK9 and LDL levels were stably reduced by 95% and 58% in mice and by 32% and 14% in macaques, respectively. ABE mRNA was cleared rapidly, and no off-target mutations in genomic DNA were found. Re-dosing in macaques did not increase editing, possibly owing to the detected humoral immune response to ABE upon treatment. These findings support further investigation of ABEs to treat patients with monogenic liver diseases.

Preparation and electroporation of Cas12a/Cpf1-guide RNA complexes for introducing large gene deletions in mouse embryonic stem cells.

CRISPR-Cas12a is a bacterial RNA-guided deoxyribonuclease that has been adopted for genetic engineering in a broad variety of organisms. Here, we describe protocols for the preparation and application of AsCas12a-guide RNA ribonucleoprotein (RNP) complexes for engineering gene deletions in mouse embryonic stem (ES) cells. We provide detailed protocols for purification of an NLS-containing AsCas12a-eGFP fusion protein, design of guide RNAs, assembly of RNP complexes, and transfection of mouse ES cells by electroporation. In addition, we present data illustrating the use of pairs of Cas12a nucleases for engineering large genetic deletions and outline experimental considerations for applications of Cas12a nucleases in ES cells.