Putting the brakes on CRISPR–Cas9

Researchers identify biological inhibitors of Cas9 nuclease activity and block genome editing in cultured human cells.

Repurposing the bacterial antiviral defense system into a genome-editing tool has transformed biomedical research. The premise of the technique, known as CRISPR–Cas9, is simple; a Cas9 nuclease is guided by an RNA molecule (sgRNA) to a complementary genome sequence that is then cut and edited. Off-target effects currently hamper CRISPR–Cas9 applications. Though there are ways to limit these unwanted effects, approaches to stopping the activity of the Cas9 nuclease have just started to emerge.

In a recent collaborative study, a team led by Alan Davidson and Karen Maxwell from the University of Toronto and Erik Sontheimer from the University of Massachusetts identified Cas9 inhibitors from the bacterium Neisseria meningitides.

The team previously described small anti-CRISPR-associated (Aca) proteins that can block type-I CRISPR systems, which rely on Cas protein complexes not typically used as genome-editing tools. Searching for Aca homologs in N. meningitides, a bacterium that possesses a Cas9-dependent type-II-C CRISPR system, the researchers have identified three new anti-CRISPRs: AcrIIC1_Nme, AcrIIC2_Nme and AcrIIC3_Nme. These proteins bind to NmeCas9 but not to Cas9 from Streptococcus pyogenes (SpyCas9), which belongs to a different Cas9 subtype. The interaction between any of these three anti-CRISPRs and NmeCas9 blocked DNA cleavage in vitro.

Forced expression of AcrIIC3_Nme together with NmeCas9 and an sgRNA in HEK293 cells inhibited gene-editing activity. Using a fluorescently tagged nuclease-dead NmeCas9 mutant, the team further showed that AcrIIC3_Nme prevented NmeCas9 or sgRNA genome binding and not DNA cleavage. However, given the great diversity in primary sequence between the three N. meningitides anti-CRISPRs, the researchers speculate that the other two inhibitors may act through different mechanisms.

Phylogenetic analysis of type-II-C Cas9 and anti-CRISPR homologs suggested that the majority of type-II-C CRISPR–Cas9 systems could be inhibited by at least one member of the three anti-CRISPR families. This encouraged the prediction that there might be at least one anti-CRISPR for any given Cas, including the SpyCas9 which is most commonly used for gene editing.

Potential applications of anti-CRISPRs are numerous and include temporal, spatial and inducible control of Cas activity, which would greatly limit unwanted gene editing.