Paper of the month November 2018

CRISPR-SKIP: Programmable gene splicing with single base editors

CRISPR/Cas9 is a programmable nuclease that can efficiently target specific locations in the genome. Typically, the nuclease induces a double stranded break (DSB) that the cell subsequently repairs either by error prone non-homologous end joining (NHEJ) or through homology directed repair (HDR). The non-dominant low-frequency HDR pathway occurs only in the S and G2 phases of the cell cycle, while the dominant NHEJ pathway provides outcomes that are highly stochastic and can lead to mutagenesis. Finally, an irreparable DSB is lethal to the targeted cell.

Single-base editors are comprised of a catalytically dead Cas9 (dCas9) fused to a cytidine deaminase and coupled to a programmable RNA guide for specificity. There are various applications for single-base editors such as protein truncation by inducing a premature stop codon as well as restoring single point mutations to their native state. By utilizing base editors, it is possible to circumvent DSBs completely and prevent stochastic outcomes by NHEJ.

In this paper, the authors utilized single-base editors to modulate alternative splicing. Alternative splicing is a naturally occurring event where the excision of introns and juxtapositioning of exons either including or excluding certain exons lead to distinct protein isoforms. Eukaryotes use alternative splicing as a sophisticated way to regulate tissue specification and development within a limited set of genes. Alternative splicing, as a therapeutic technique, excludes exons that contain malignant errors paving the way for healthy isoforms expression. When applied, this method can determine the different functions that distinct protein isoforms have in tissue specification and development. Furthermore, some transcripts, such as long non-coding RNAs (lncRNAs), cannot be knocked-out using traditional methods. However, modification via alternative splicing allows their regulatory role to be determined.  

The authors hypothesized that targeting the highly conserved guanine in the splice acceptor region between exons and introns would disrupt the recognition of the spliceosome and as such, exclude the following exon from the mature transcript. They selected genes based on splice sites in the inner exons that were in range of appropriate protospacer adjacent motifs (PAMs) with a multiple of three to avoid possible frameshift. They estimated that 118 089 inner exons in the human genome fulfill the criteria making the method highly dynamic. Various cell lines from both human and murine origin were transfected with guides targeting several genes, mainly involved in cancer metabolism.  The Cas9 proteins used were SaCas9-KKH-BE3, SpCas9-BE3, SpCas9-VRER-BE3, SpCas9-VQR-BE3 with the primary PAMs NNNRRT, NGG, NGCG and NGA, respectively. By using different Cas9 proteins that recognized different PAMs, they were able to increase the amount of genes targeted. They found that although there was significant base editing at 77.78% of targets, induced exon skipping occurred in only 50% of targets. A possible explanation for this might be the existence of cryptic splice sites that made exon inclusion possible.

Although the authors managed to get significant exon skipping, the method is flawed. In addition to cryptic splice sites, one main limitation is the need of a PAM within 12-17 bp from the target base. However, by using additional Cas9 proteins the system becomes more flexible.  In addition, there were some off target events. However, due to emerging high fidelity base editors that require a longer RNA-DNA complementarity, such events could decrease and be a valuable addition to the technique.

The results demonstrate that single-base editors used significantly induce exon exclusion and thus, be a putative therapeutic technique for diseases with pathological protein isoforms. When applied, the technique can also elucidate the function of lncRNAs, which would otherwise be challenging due to the fact that they are non-protein coding and cannot be knocked out through frameshift mutations.  By discovery of additional Cas9 proteins and higher fidelity base editors, the technique could improve and be used in gene therapy.