Shorter Guide Strands Increase CRISPR/Cas Nuclease Specificity
May 2014
Clustered, regularly interspaced, short palindromic repeat (CRISPR) RNA-guided nucleases have recently gained a lot of attention for their ability to manipulate the genome. These nucleases are particularly useful because they can target a specific genomic location through the use of guide RNA (gRNA). In nature, gRNA is approximately 100 nucleotides long. The specificity is determined by a 20 bp sequence ending with an NGG trinucleotide (5′-X(20)NGG-3′) that binds to the target RNA to be edited (Seung Woo Cho et al., 2013). gRNA tolerates up to 5 mismatches, particularly at the 5′ region. This can introduce unwanted insertion or deletion (indel) mutations, limiting the use of the CRISPR/Cas9 system. To combat this problem, Fu et al. attempt to improve the specificity through manipulation of the guide sequences in their manuscript entitled, Improving CRISPR-Cas nuclease specificity using truncated guide RNAs.
The authors hypothesize that a reduction in the length of the recognition sequence (located at the gRNA/DNA interface) will increase specificity. First, they evaluated if a shorter recognition sequence would recognize and target efficiently. To do this, the authors utilized a disruption assay where efficient targeting introduces an indel mutation and inhibits translation. They constructed gRNA with 15, 17, 19 or 20 complementary nucleotides to a region of eGFP and found that truncated gRNA (tru-gRNA) of 17 and 19 bases targeted as efficiently as those with 20 bases. In contrast, tru-gRNA with 15 complementary nucleotides did not show detectable levels of targeting. Next they utilized the same assay to assess the generality of this finding by constructing tru-gRNA to four different sites in eGFP and found that various tru-gRNA with a minimum of 17 complementary nucleotides were able to target all sites efficiently. Next, the authors constructed gRNA of 17 to 20 bases to endogenous human genes and assayed for insertions or deletions at the corresponding region. They found comparable targeting efficiency of the tru-gRNA as compared to full-length gRNA at 5 of the 7 sites. The authors note that while the tru-gRNA efficiency was reduced at two of the tested sites, the absolute rate of mutagenesis remained high (13.3% and 16.6%). Next Fu and colleagues investigated sensitivity by introducing single or double Watson-Crick mismatches within the gRNA at various sites within the gRNA/DNA interface. Using the GFP disruption assay, they found that tru-gRNA was more sensitive than full-length gRNA, though the degree of sensitivity was site dependent.
As a more sensitive measure, the authors then utilized a T7E1 assay to assess off-target mutation rates of tru-gRNA and full-length gRNA targeted to three sites on two human genes. These sites were chosen because previous studies already reported off-target mutation sites with full-length gRNA. They found that the tru-gRNA decreased the rate of mutation at all tested off-target sites and in some cases it was decreased as much as 5000 fold, compared to full-length gRNA. Three sites showed no detectable indels with the tru-gRNA. To test whether tru-gRNA could mutate sites not previously reported, the authors identified 97 additional potential off-target sites (1-3 mismatches) in two different cell lines. They reported no detectable indel mutations in the human U2OS.EGFP cell line and only one in human FT-HEK293 cells with the tru-gRNA. Full-length gRNA was shown to also induce a mutation at this site. This suggests that tru-gRNA is not likely to introduce more mutations than full-length gRNA.
Finally, because the T7E1 assay may not detect very low levels of mutation, the authors took the 30 potential off-target sites that contained the fewest mismatches and deep sequenced those regions. They found off-target indel mutations caused by the tru-gRNA to be undetectable or at very low levels (0.00-0.35%). Interestingly, when they tested whether tru-gRNA would elicit similar favorable results with Cas9 nickase (D10A), they found the lowest levels of mutation when they utilized one tru-gRNA combined with one full-length gRNA (0.00%-0.001%).
The development of the CRISPR/Cas9 technology has improved the ease of gene manipulation in many systems. Off-target effects threatened to be a primary roadblock in the use of this technology. However, Fu et. al. convincingly demonstrate that truncation of the recognition sequence can reduce the off-target mutation rate thus improving current methods. This finding may prove to be vital for the advancement of the CRISPR technology as a therapeutic tool. TriLink offers custom synthesis of gRNA as well as Cas9 mRNA. The use of our products together avoids the use of plasmids or viral vectors to eliminate the risk of insertion of plasmid DNA at the cut site. This free resource is available for guidance on designing full-length or tru-gRNA.
Have a question? Visit Ask An Expert.