引导RNA
Cas9
DNA
清脆的
化学
核糖核酸
计算生物学
生物
生物化学
基因
作者
Minjie Kang,Zhicheng Zuo,Zhixiang Yin,Jun Gu
标识
DOI:10.1021/acs.jcim.1c01562
摘要
The off-target effects of Streptococcus pyogenes Cas9 (SpCas9) pose a significant challenge to harness it as a therapeutical approach. Two major factors can result in SpCas9 off-targeting: tolerance to target DNA–guide RNA (gRNA) mismatch and less stringent recognition of protospacer adjacent motif (PAM) flanking the target DNA. Despite the abundance of engineered SpCas9-gRNA variants with improved sensitivity to target DNA–gRNA mismatch, studies focusing on enhancing SpCas9 PAM recognition stringency are quite few. A recent pioneering study identified a D1135E variant of SpCas9 that exhibits much-reduced editing activity at the noncanonical NAG/NGA PAM sites while preserving robust on-target activity at the canonical NGG-flanking sites (N is any nucleobase). Herein, we aim to clarify the molecular mechanism by which this single D1135E mutation confers on SpCas9 enhanced specificity for PAM recognition by molecular dynamics simulations. The results suggest that the variant maintains the base-specific recognition for the canonical NGG PAM via four hydrogen bonds, akin to that in the wild type (WT) SpCas9. While the noncanonical NAG PAM is engaged to the two PAM-interacting arginine residues (i.e., R1333 and R1335) in WT SpCas9 via two to three hydrogen bonds, the D1135E variant prefers to establish two hydrogen bonds with the PAM bases, accounting for its minimal editing activity on the off-target sites with an NAG PAM. The impaired NAG recognition by D1135E SpCas9 results from the PAM duplex displacement such that the hydrogen bond of R1333 to the second PAM base is disfavored. We further propose a mechanistic model to delineate how the mutation perturbs the noncanonical PAM recognition. We anticipate that the mechanistic knowledge could be leveraged for continuous optimization of SpCas9 PAM recognition specificity toward high-precision demanding applications.
科研通智能强力驱动
Strongly Powered by AbleSci AI