Slaymaker et al (2015) describe in Science their successful efforts to modify Cas9 so that it retains its on-target activity but has highly reduced off-target activity.
The potential bane of a well-designed and executed gene editing experiment are off-target mutations that confound or obscure the phenotypic effects of the mutated target gene. This is a familiar problem since it is also encountered when using dsRNA for gene silencing. Mitigating these potential problems is difficult and the ‘acid’ test for showing that your targeted mutation is indeed responsible for the observed phenotype is to revert or complement the mutation and see if a wild-type phenotype is restored. This can be difficult or impractical so often the best one can do is to mitigate possible off target effects and there are a number of good options for doing this that the authors review.
Slaymaker et al. have potentially provided researchers with an excellent simple means to mitigate off target effects by creating variants of Cas9 with reduced off-target activity but with ‘normal’ on-target activity. They did this using some clever protein engineering based on the known structure and mechanism by which Cas9 interacts with and cuts DNA.
Cas9 cuts DNA following strand separation and R-loop formation. The authors reasoned that if they could prevent strand separation when the guide RNA-loaded Cas9 was interacting with an off-target site they could reduce off-target site cleavage. The authors proposed ” that off-target cutting occurs when the strength of Cas9 binding to the non-target DNA strand exceeds forces of DNA rehybridization.”
So, their approach was to mutate Cas9 (neutralize some positive charges) so that its ability to bind to the non-target DNA strand was reduced. The location in Cas9 where the non-target strand of DNA interacts was known, giving the investigators a good region to mutagenize. They created 32 Cas9 variants with a single amino acid change (to Alanine) and found that some had the desired characteristics. They made variants that contained multiple mutations and found 3 that were of particular interest.
They tested these using validated guide RNAs to known targets and with known and characterized off-target sites. The 3 variants retained target specificity and lost off-target activity. In some cases there was no detectable off-target activity. In the end and after a variety of other assessments that can be found in the published paper, one variant emerged as the most interesting – SpCas9 (K848A/K1003A/R1060A) also referred to as eSpCas9(1.1).
By neutralizing some positive charges and weakening the interactions of Cas9 with the non-target DNA strand off target cleavage was drastically reduced while retaining on-target activity and specificity. eSpCas9(1.1) is available to researchers and more information about this system is available on the senior author’s website www.genome-engineering.org
It is likely this will be the reagent of choice in the near term although one can not forget about off-target effects.
Ian M. Slaymaker, Linyi Gao, Bernd Zetsche, David A. Scott, Winston X. Yan, and Feng Zhang (2015) Rationally engineered Cas9 nucleases with improved specificity Science aad5227 Published online 1 December 2015 [DOI:10.1126/science.aad5227]