A CRISPRi screen in E. coli reveals sequence-specific toxicity of dCas9

Abstract

High-throughput CRISPR-Cas9 screens have recently emerged as powerful tools to decipher gene functions and genetic interactions. Here we use a genome-wide library of guide RNAs to direct the catalytically dead Cas9 (dCas9) to block gene transcription in Escherichia coli. Using a machine-learning approach, we reveal that guide RNAs sharing specific 5-nucleotide seed sequences can produce strong fitness defects or even kill E. coli regardless of the other 15 nucleotides of guide sequence. This effect occurs at high dCas9 concentrations and can be alleviated by tuning the expression of dCas9 while maintaining strong on-target repression. Our results also highlight the fact that off-targets with as little as nine nucleotides of homology to the guide RNA can strongly block gene expression. Altogether this study provides important design rules to safely use dCas9 in E. coli.

Pooled CRISPRi screening of the cyanobacteriumSynechocystissp. PCC 6803 for enhanced growth, tolerance, and chemical production

“…There is also a 75% agreement in our calculated fitness scores for genes in central carbon metabolism to their orthologs in Synechoccocus PCC 7942, the latter determined by Tn-Seq 14 (Supplemental Data 10). Third, off-target binding by dCas9 was shown to be problematic for strongly expressed dCas9 61 , while our genome-integrated dCas9 was driven by a weak promoter 62 . However, dCas9 is potent, as we observed gene repression for some targets even in the absence of the inducer (Cluster 1 in Fig.…”

Strategies for Developing CRISPR‐Based Gene Editing Methods in Bacteria

“…The heteroduplex of sgRNA and the target DNA strand stabilizes when approximate eight consecutive PAM‐proximal matches (seed region), regardless of mismatches present in the PAM distal end . This phenomenon also indicated the reason why off‐target bindings frequently occurred in dSpCas9 . However, the extent of the heteroduplex toward the PAM‐distal region determines the activity of DNA cleavage .…”

“…Generally, dCas9‐mediated DNA binding requires fewer base pairings between sgRNA and target DNA than Cas9‐led cleavage, since dCas9 do not require complete binding over seed region to trigger the conformational change, which leads to DNA cleavage . Cui et al demonstrated that when dCas9 was expressed constitutively at a high level in E. coli , sgRNAs with specific 5‐bp seed sequences at the PAM‐proximal region would produce substantial fitness defect, even when the other 15‐bp in sgRNAs were completely different or the targeted gene was nonessential. These 5‐bp seed sequences have been characterized as 5′‐ACCCA‐3′, 5′‐ATACT‐3′, and 5′‐TGGAA‐3′, and designated as “bad seed.” The authors found that this effect could be alleviated by reducing the expression level of dCas9.…”

Impact of CRISPR interference on strain development in biotechnology

“…A more general approach was followed when the dCas9 enzyme was engineered to reduce its toxicity, for example, by eliminating PAM binding with a R1335K mutation (dCas9*) and recovering DNA binding by fusing dCas9 to the PhlF repressor (dCas9*_PhlF) in E. coli [147]. One reason for dCas9 toxicity was described as “bad seed” effect, that is, unexplained sequence‐specific toxicity of the chosen PAM and gRNA sequences, which led to dCas9 overexpression [138,148]. A machine‐learning approach revealed that gRNAs sharing specific 5‐nucleotide seed sequences can produce strong fitness defects or even kill E. coli regardless of the other 15 nucleotides of guide sequence [138].…”

“…One reason for dCas9 toxicity was described as “bad seed” effect, that is, unexplained sequence‐specific toxicity of the chosen PAM and gRNA sequences, which led to dCas9 overexpression [138,148]. A machine‐learning approach revealed that gRNAs sharing specific 5‐nucleotide seed sequences can produce strong fitness defects or even kill E. coli regardless of the other 15 nucleotides of guide sequence [138]. Moreover, the off‐target effects of dCas9 have been reported to increase with the number of off‐target binding sites depending on the constructed gRNA [142,143,149].…”