Profiling single-guide RNA specificity reveals a mismatch sensitive core sequence

Ting Zheng,Jianxin Chen,York Yuanyuan Zhu,Da Liang,Yi Yang,Ying Xu,Wei Peng,Pingjing Zhang,Leilei Niu,Li-He Zhang,Yingzi Hou,Fan Yi,Wenjian Feng,Quan Du,Zhenxi Zhang,Ying Yang,Letian Zhang

doi:10.1038/srep40638

Abstract

Targeting specificity is an essential issue in the development of CRISPR-Cas technology. Using a luciferase activation assay, off-target cleavage activity of sgRNA was systematically investigated on single nucleotide-mismatched targets. In addition to confirming that PAM-proximal mismatches are less tolerated than PAM-distal mismatches, our study further identified a “core” sequence that is highly sensitive to target-mismatch. This sequence is of 4-nucleotide long, located at +4 to +7 position upstream of PAM, and positioned in a steric restriction region when assembled into Cas9 endonuclease. Our study also found that, single or multiple target mismatches at this region abolished off-target cleavage mediated by active sgRNAs, thus proposing a principle for gene-specific sgRNA design. Characterization of a mismatch sensitive “core” sequence not only enhances our understanding of how this elegant system functions, but also facilitates our efforts to improve targeting specificity of a sgRNA.

Highlights

With the development of CRISPR-Cas[9] system, researchers are able to induce specific double-stranded breaks in chromosomal DNA, whose repair leads to either local mutations or homologous recombination with donor DNA
Utility of CRISPR-Cas system relies on accurate selection of target DNA, in which process three components are engaged, including Cas[9] protein, stranded guide RNA (sgRNA) and target DNA
Correct target selection derives from the base-pairing between a 20-nt sgRNA sequence and the DNA, as well as the presence of a protospacer adjacent motif (PAM) sequence

Summary

Results and Discussion

Utility of CRISPR-Cas system relies on accurate selection of target DNA, in which process three components are engaged, including Cas[9] protein, sgRNA and target DNA. While the cleavage efficacy varied with the position and the nucleotide identity in the seed region depending on the sgRNA sequence, the most profound compromising effects were surprisingly observed within a short “core” sequence, ranging from positions +4 to +7 (Fig. 3). Target cleavage was abolished by most of the single-nucleotide mismatches at this “core” sequence, rendering Cas[9] activity highly sensitive to the mismatches This suggests a strict requirement of an intact A-form architecture for this region, which might attribute to its spatial restriction within Cas[9]. Our data indicate that target specificity of a sgRNA may be determined by a variety of mechanisms, depending on its interactions with target DNA, neighbouring protein domains, and solvent molecules This is of notable for a “core” sequence within the PAM-proximal region. To maximize target specificity of a sgRNA, we suggest that potential ‘off-target’ genomic sites should be examined by considering the following guidelines: (1) potential ‘off-target’ sites should not be followed by a NGG or NAG PAM sequences; (2) global sequence similarity between sgRNA and potential ‘off-target’ sites should be minimized; (3) single or multiple target mismatch in the “core” sequence is preferred

Materials and Methods

Author Contributions

Additional Information