A bacterial enzyme that uses guide RNA molecules to target DNA for cleavage has been adopted as a programmable tool to site-specifically modify genomes of cells and organisms, from bacteria and human cells to whole zebrafish.
In a 1987 paper, researchers at Osaka University in Japan reported an apparently minor finding. While investigating the sequence of a bacterial gene that encodes the enzyme alkaline phosphatase, they discovered an unusual segment of neighbouring DNA that consisted of short, directly repeating nucleotide sequences flanked by short unique segments1. They noted that “the biological significance of these sequences is not known”. Fast-forward almost three decades, and what initially seemed to be an obscure observation is now being used to open the door to easy manipulation of the genomes of a multitude of organisms. Five papers published within a month of each other, in Science2,3 and Nature Biotechnology4,5,6, report the application of such bacterial sequences — now referred to as CRISPR–Cas systems7 — as a simple and versatile tool for genomic editing.
As whole-genome sequencing became routine in recent decades, regions containing CRISPR (clustered regularly interspaced short palindromic repeat) sequences and CRISPR-associated (Cas) genes were found in a wide variety of bacteria and archaea8,9,10,11,12. The discovery10,11 that the short unique sequences in these arrays matched DNA sequences from viruses or plasmids (small non-chromosomal DNA molecules that can be transferred among bacteria and archaea) hinted that CRISPR–Cas systems encode 'adaptive' immune systems, providing specific defences against invaders. Subsequent genetic and biochemical experiments confirmed this speculation by showing that CRISPR–Cas systems allow detection of and protection against mobile genetic elements13.
Although some CRISPR–Cas systems require multiple proteins to function14, the type II systems found in many bacteria13,15,16 use a single endonuclease, Cas9 (Fig. 1). This enzyme acts together with guide RNA to locate and cleave invading DNA at sites demarcated by conserved sequences called proto-spacer adjacent motifs (PAMs)7,17,18. To form a functional DNA-targeting complex, Cas9 requires two distinct RNA transcripts, CRISPR RNA (crRNA) and trans-acting CRISPR RNA (tracrRNA)7,15. However, it was recently shown7 that this dual RNA can be reconfigured as a single-guide RNA (sgRNA) that includes sequences that are sufficient to program Cas9 to introduce double-stranded breaks in target DNA. As the new publications show, RNA-guided Cas9 can function in a variety of cells and organisms to cleave intact genomes at specific sites. And this is the point at which the potential for genome editing comes in. When the double-stranded breaks are repaired by standard cellular repair mechanisms, either by homologous recombination (the exchange of genetic information between DNA molecules with similar sequences) or non-homologous end joining (NHEJ; the introduction of insertions or deletions into the sequence), the sequence at the repair site can be modified or new genetic information inserted.
Three of the studies demonstrate that RNA-programmed Cas9 can function in human cells. Cong et al.2, Mali et al.3 and Cho et al.4 engineered versions of the Cas9 enzyme from the bacterium Streptococcus pyogenes so that it would be active in the nucleus of human cells, and designed dual RNAs or sgRNAs that included a 20-nucleotide sequence complementary to human target DNA sequences. When the researchers expressed the 'humanized' Cas9 together with these guide RNAs in various human cell lines, including induced pluripotent stem cells, they observed the expected alterations to the target DNA — achieved through the introduction of double-stranded breaks followed by repair. The gene-targeting achieved up to 38% success and was accompanied by only a low level of Cas9 toxicity. The RNA-guided Cas9 was also efficient at triggering targeted gene replacement at normal genomic sites in human cells. In another paper published in the same month, Jinek et al.19 show that RNA-programmed Cas9 functions in human cells to trigger site-specific genome modifications, and that the ability of Cas9 to assemble with guide RNA in cells is a limiting factor in this activity.
On the basis of earlier observations that single-stranded DNA breaks can favour homologous recombination and reduce off-target mutagenesis, Mali et al.3 and Cong et al.2 also tested versions of Cas9 that have been shown7 to act as a nickase enzyme — one that breaks only one strand of a DNA molecule. The mutated enzymes had lower rates of NHEJ but were as efficient as the wild-type endonuclease at gene replacement triggered by homologous recombination. Both groups also demonstrated further functionality of the system in 'multiplexed' targeting; the expression of sgRNA-programmed Cas9 that can bind to two different genomic sequences led to sequence disruption at more than one independent target site. In addition, Cong and colleagues show that gene-disruption efficiency could be improved upon independent expression of the two RNA components of the original dual-tracrRNA–crRNA combination. This finding implies that improved design of sgRNAs may allow them to better mimic the dual RNA structure7.
In addition to these results in cell lines, RNA-guided Cas9 can be used to engineer genomic changes in intact organisms. Jiang and colleagues6 show that the system can be used in bacteria to modify multiple sites by programming Cas9 with several different guide RNAs in a single cell. This technology could be exploited to engineer microorganisms that are otherwise genetically intractable to harbour pathways for producing biofuels and molecules of therapeutic value. Working with zebrafish, Hwang and colleagues5 show that injection of one-cell-stage embryos with Cas9-encoding mRNA and appropriate guide RNAs produced high frequencies (24–59%) of targeted insertions and deletions at eight of ten sites in all embryos tested. These findings hint that RNA-guided Cas9 might be useful for engineering other multicellular organisms, including mammals and plants. One of the most exciting potential uses of such technology would be to provide a straightforward means of generating animal models of human disease.
Genome engineering by RNA-programmable Cas9 promises to have broad applications in synthetic biology, direct and multiplexed perturbation of gene networks, and targeted ex vivo and in vivo gene therapy2,3,4,5,6,7. The next challenges will be to analyse and address possible off-target effects and improve the efficiency and specificity of the system, while expanding its use to other organisms. In this regard, it will be important to compare RNA-programmed Cas9 with existing genome-editing tools18, including meganucleases, ZFNs (zinc-finger nucleases) and TALENs (transcription activator-like effector nucleases). In addition to genome editing, this approach offers the exciting possibilities of transcriptional gene silencing using an inactive Cas9 (ref. 20) or engineering Cas9 to have new functions, such as transcriptional activation. The discovery and application of bacterial systems, such as restriction enzymes and thermostable polymerases, have revolutionized molecular biology in the past. With RNA-guided Cas9 enzymes, bacteria now offer a versatile tool for rewriting genomic sequence information that has the potential to reshape the genome-engineering landscape in biotechnology and medicine.
About this article
Annals of the New York Academy of Sciences (2019)