A CRISPR-Cas9 Assisted Non-Homologous End-Joining Strategy for One-step Engineering of Bacterial Genome

Abstract

Homologous recombination-mediated genome engineering has been broadly applied in prokaryotes with high efficiency and accuracy. However, this method is limited in realizing larger-scale genome editing with numerous genes or large DNA fragments because of the relatively complicated procedure for DNA editing template construction. Here, we describe a CRISPR-Cas9 assisted non-homologous end-joining (CA-NHEJ) strategy for the rapid and efficient inactivation of bacterial gene (s) in a homologous recombination-independent manner and without the use of selective marker. Our study show that CA-NHEJ can be used to delete large chromosomal DNA fragments in a single step that does not require homologous DNA template. It is thus a novel and powerful tool for bacterial genomes reducing and possesses the potential for accelerating the genome evolution.

Figure 1. One-step inactivation of chromosomal gene(s) by CRISPR-Cas9 assisted non-homologous end-joining (CA-NHEJ).

Cas9 and NHEJ-related proteins (Mt-Ku and Mt-LigD) are expressed in host cells, which are then transformed with a single-guide RNA (sgRNA) donor plasmid to generate double-stranded breaks (DSBs) and trigger indel mutations. Mutagenesis is attributed to the RNA-directed Cas9 cleavage system and the error-prone NHEJ repair system. First, site-specific DSB is generated via sgRNA-directed Cas9 cleavage. The DNA ends are recognized and stabilized by the DNA end-binding protein Mt-Ku. Next, the ATP-dependent DNA ligase Mt-LigD is recruited to the DNA ends; the imprecise repair of DSB results in a frameshift mutation. Finally, only the DSB-repaired colonies lacking the Cas9 targeting site survive CRISPR-Cas9 screening. To further engineer the strain, the sgRNA donor plasmid is cured via an inducible sgRNA-mediated “suicide” strategy, and the temperature-sensitive plasmid pCas9 (Ts)-NHEJ by growing the cells at 42 °C.

Figure 2. Efficiency and mutation rates of the hetereologous non-homologous end-joining (NHEJ) pathway in E. coli.

(a) Efficiency of the NHEJ system in re-circularizing the in vitro HindIII- or SmaI-digested pUC19 plasmid and the in vivo CRISPR-Cas9 cleaved pUC-lacZ plasmid. The error bars represent standard deviations from three replicate experiments. The results are expressed as colony-forming units (CFU) per μg of plasmid DNA. mku and ligd, derived from M. tuberculosis H37Rv and involved in the NHEJ pathway, are hetereologously expressed in E. coli using a strong constitutive P_J23119 promoter. (b) The mutation rates of the NHEJ system with different artificially created DSBs. The mutation rates are statistically determined based on the proportion of white colonies on the X-gal plate. The error bars represent standard deviations from three replicate experiments.

Figure 3. Gene disruption using CA-NHEJ in E. coli.

(a) The structure of the natural CRISPR array and the artificial single-guide RNA (sgRNA). White: repeat sequences; red: spacer sequences; blue: RNA domains for Cas9 binding and a 40-nt transcription terminator. (b) The lacZ mutagenesis positivity rate in strain MG1655 using spacers expressed by the CRISPR array or sgRNA cassette. Two spacers, L4 and LR4, targeting the sense strand and the antisense strand of lacZ, were tested, respectively. The positivity rate was statistically calculated based on the proportion of white colonies on the X-gal plate. The error bars represent standard deviations from three replicate experiments. (c) Distribution of sgRNAs designed to target the lacZ gene in strain MG1655. The location of the designed spacers, the spacer sequences, the PAM (protospacer adjacent motif) sequences, and the cleavage sites are detailed and highlighted. (d) The positivity rate of lacZ mutagenesis and the range of DNA end fragment deletion using various sgRNAs. The positivity rate shown is representative of three replicate experiments. Solid square: the positivity rate of lacZ mutagenesis; dark cyan bar: the length of DNA end fragment deletion. For each sgRNA target, eight white colonies were randomly picked for Sanger sequencing to determine the length of the deleted fragment at the junction.

Figure 4. Large DNA fragment deletion in the E. coli MG1655 chromosome by CA-NHEJ.

(a) Schematic of chromosomal fragment deletions using CA-NHEJ. Four sgRNA pairs, L4&LR8, LI10&LA0, LI10&CR0, ME17&CR0, were designed for the deletion of the lacZ gene, lac operon, lac-cyn operons, and mhp-lac-cyn operons, respectively. The distribution of the sgRNA pairs and the DNA fragments expected to be deleted by separate sgRNA pairs are detailed and highlighted. Cyan: the mhp operon; tan: the lac operon; green: the cyn operon. (b) PCR analysis of six transformants from each engineering experiment, used to identify the lengths of deleted DNA fragments by various sgRNA pairs.