Credit: ImageSource

Genome-editing nucleases are valuable for understanding gene function and have potential gene therapy applications. An endogenous sequence-specific nuclease system is the bacterial and archaeal clustered regularly interspaced short palindromic repeats (CRISPR) system, which uses genomically encoded RNA arrays to guide the cleavage of foreign DNA by a CRISPR-associated (Cas) nuclease. Two new studies have adapted this system to achieve genome editing in mammalian cells.

this system is amenable to multiplexing

Mali et al. and Cong et al. made mammalian expression constructs from the Streptococcus pyogenes CRISPR components, including nuclear-localized Cas9 and processed forms of guide RNAs. Both teams showed that transfection of their constructs into various human cell types resulted in the characteristic disruptions of intended target genes that are expected from the repair of Cas9-generated DNA double-strand breaks (DSBs) by the mutagenic non-homologous end joining (NHEJ) machinery. The efficiencies were variable depending on the cell type and construct design, but up to 38% of targeted alleles in the cell population were disrupted.

Interestingly, although both groups generally expressed a single guide RNA construct that was a chimaera of an engineered guide sequence and the common RNA sequence that is required for Cas9 interaction, Cong et al. also showed that expressing the two RNA components separately (as occurs in the endogenous CRISPR system) can improve gene disruption efficiency. Moreover, both groups showed that encoding two target sequences in the guide RNA resulted in disruption at both genomic sites, implying that this system is amenable to multiplexing.

In addition to initiating targeted gene disruptions, genome editing can also take advantage of the cellular homologous recombination machinery to replace a DNA locus with an engineered donor DNA casette that contains regions of homology flanking the DSB site. Mali et al. used the CRISPR system to repair a genomically integrated but defective GFP construct in human cells at 8% efficiency, as shown by the detection of fluorescence by flow cytometry. Furthermore, both groups achieved DNA replacement at endogenous human genomic loci.

One way to promote homologous recombination and to minimize NHEJ-mediated off-target mutagenesis is to use a mutant nuclease (a 'nickase') that generates only DNA single-strand breaks. Both groups showed that a nickase form of Cas9 was equally active for homologous-recombination-mediated locus replacement but generated fewer NHEJ-mediated insertions and deletions.

Overall, both teams reported that the CRISPR system has comparable or superior genome-editing efficiency compared with systems based on transcription-activator-like effector nucleases (TALENs) or zinc finger nucleases (ZFNs), although it will be important fully to assess the relative potency and site-specificity of these methods. It will also be interesting to determine the opportunities and challenges in tool design that RNA-based guiding of CRISPR systems provides compared with the protein-mediated TALENs and ZFNs.