Guide-target mismatch effects on dCas9–sgRNA binding activity in living bacterial cells

• Published in: Nucleic acids research Vol. 49; no. 3; pp. 1263 - 1277
• Main Authors: Feng, Huibao, Guo, Jiahui, Wang, Tianmin, Zhang, Chong, Xing, Xin-hui
• Format: Journal Article
• Language: English
• Published: England Oxford University Press 22.02.2021
• Subjects: Base Pair Mismatch; Chemical Biology and Nucleic Acid Chemistry; CRISPR-Associated Protein 9 - metabolism; CRISPR-Cas Systems; DNA - metabolism; Escherichia coli - genetics; Models, Genetic; Protein Binding; RNA - chemistry; RNA - metabolism; Thermodynamics
• ISSN: 0305-1048; 1362-4962
• DOI: 10.1093/nar/gkaa1295

Abstract As an effective programmable DNA targeting tool, CRISPR–Cas9 system has been adopted in varieties of biotechnological applications. However, the off-target effects, derived from the tolerance towards guide-target mismatches, are regarded as the major problems in engineering CRISPR systems. To understand this, we constructed two sgRNA libraries carrying saturated single- and double-nucleotide mismatches in living bacteria cells, and profiled the comprehensive landscape of in vivo binding affinity of dCas9 toward DNA target guided by each individual sgRNA with particular mismatches. We observed a synergistic effect in seed, where combinatorial double mutations caused more severe activity loss compared with the two corresponding single mutations. Moreover, we found that a particular mismatch type, dDrG (D = A, T, G), only showed moderate impairment on binding. To quantitatively understand the causal relationship between mismatch and binding behaviour of dCas9, we further established a biophysical model, and found that the thermodynamic properties of base-pairing coupled with strand invasion process, to a large extent, can account for the observed mismatch-activity landscape. Finally, we repurposed this model, together with a convolutional neural network constructed based on the same mechanism, as a predictive tool to guide the rational design of sgRNA in bacterial CRISPR interference.