Kleinstiver et al. describe in Nature their successful efforts to create high-fidelity SpCas9 nuclease variants that have high on-target efficiencies and low or no off-target effects. The results are SpCas9-HF1.
When using Cas9 or other engineered nuclease systems one hopes for efficient on-target activity to create the cuts or nicks in the DNA sequence of interest. Of course, one also worries about whether the nuclease acted at other sites within the genome – either because there were other perfect matches to the target site of interest or because there were sequences that resembled the target site sufficiently to allow Cas9/gRNA interaction and nuclease activity. Off-target activity is challenging to detect and is a constant concern.
Eliminating off-target mutations is a major challenge and a number of strategies have emerged that are useful.
Kleinstiver et al. hypothesized that if non-specific interactions between SpCas9 and the target DNA site could be reduced then specificity could be increased. They proposed that the SpCas9/gRNA complex had more than enough energy for optimal recognition of a target site and consequently could tolerate mismatches. So, they reasoned, if they could reduce this ‘excess energy’ they could keep the specificity of the complex while reducing off target activity because the complex would no longer be able to stably associate with imperfect targets.
There are some specific amino acids in SpCas9 that make contact with the DNA target. They identified and then mutagenized 4 amino acid positions and various combinations of these mutants and then assessed the efficiency and precision of the resulting SpCas9 proteins using a robust assay performed in human cell lines.
An SpCas9 variant with four substitutions at these contact sites had 70% of the on-target activity of wild-type SpCas9 but had no detectable off-target activity as revealed by their GUIDE-seq analysis – a method for recovering and identifying genomic DNA that has been cut by SpCas9 (in principal the method resembles methods like RAD-seq). This variant was called SpCas9-HF1 and had substitutions N497A, R661A, Q695 and Q926A.
The authors describe their rather extensive efforts to evaluate SpCas9-HF1 and come to the conclusion that this variant can “completely or nearly completely reduce off-target mutations …to levels that are undetectable by GUIDE-seq and targeted deep sequencing.” (quoted from paper).
They also show that SpCas9-HF1 can be used to target simple repeat sequences and targets containing homopolymeric seuquencences.
Not long ago Slaymaker et al. (2015) also described a high-fidelity variant of SpCas9 that was constructed using some of the same logic applied by Kleinstiver et al. The “Slaymaker variant”, eSpCas9(1.1), contained three substitutions – K848A, K1003A, R1060A.
SpCas9-HF1, like eSpCas9(1.1), is freely available through Addgene.
Kleinstiver, B. P., Pattanayak, V., Prew, M. S., Tsai, S. Q., Nguyen, N. T. et al., 2016 High-fidelity CRISPR–Cas9 nucleases with no detectable genome-wide off-target effects. Nature advance online publication: doi: 10.1038/nature16526
Recent Related Posts