A CRISPR-Assisted Nonhomologous End-Joining Strategy for Efficient Genome Editing in Mycobacterium tuberculosis

Abstract

New tools for genetic manipulation of Mycobacterium tuberculosis are needed for the development of new drug regimens and vaccines aimed at curing tuberculosis infections. Clustered regularly interspaced short palindromic repeat (CRISPR)-CRISPR-associated protein (Cas) systems generate a highly specific double-strand break at the target site that can be repaired via nonhomologous end joining (NHEJ), resulting in the desired genome alteration. In this study, we first improved the NHEJ repair pathway and developed a CRISPR-Cas-mediated genome-editing method that allowed us to generate markerless deletion in Mycobacterium smegmatis, Mycobacterium marinum, and M. tuberculosis Then, we demonstrated that this system could efficiently achieve simultaneous generation of double mutations and large-scale genetic mutations in M. tuberculosis Finally, we showed that the strategy we developed can also be used to facilitate genome editing in Escherichia coli IMPORTANCE The global health impact of M. tuberculosis necessitates the development of new genetic tools for its manipulation, to facilitate the identification and characterization of novel drug targets and vaccine candidates. Clustered regularly interspaced short palindromic repeat (CRISPR)-CRISPR-associated protein (Cas) genome editing has proven to be a powerful genetic tool in various organisms; to date, however, attempts to use this approach in M. tuberculosis have failed. Here, we describe a genome-editing tool based on CRISPR cleavage and the nonhomologous end-joining (NHEJ) repair pathway that can efficiently generate deletion mutants in M. tuberculosis More importantly, this system can generate simultaneous double mutations and large-scale genetic mutations in this species. We anticipate that this CRISPR-NHEJ-assisted genome-editing system will be broadly useful for research on mycobacteria, vaccine development, and drug target profiling.

Keywords: CRISPR-Cas system; Mycobacterium marinum; Mycobacterium smegmatis; Mycobacterium tuberculosis; genome editing; nonhomologous end joining.

FIG 1

Cartoon representation of CRISPR-Cas–NHEJ-assisted genome editing in mycobacteria. Genome editing is completed in two steps: (i) cleavage by the CRISPR-Cas system and (ii) NHEJ-mediated repair. High expression of the MmNHEJ machinery (Ku, LigD, and NrgA from M. marinum) facilitates efficient genome editing in M. marinum (bottom left, “With RecA”). Repression of RecA-dependent homologous recombination (HR) can increase the efficiency of NHEJ, thereby increasing genome-editing efficiency in M. smegmatis (bottom left, “Repression of RecA”). NHEJ efficiency is higher when double-strand breaks are generated during stationary phase; thus, efficient genome editing can be achieved in RecA-repressed M. tuberculosis cells in stationary phase (bottom right).

FIG 2

CRISPR-Cas12a-assisted NHEJ genome editing in M. marinum. (A) Genome-editing efficiency in the presence of CRISPR-Cas12a cleavage in M. marinum. Plasmids pYC1103, pYC1178, and pYC1521, expressing whiB6, recD, and nrgA-targeting crRNA, respectively, were transformed into M. marinum with the Cas12a-expressing plasmid pMV261-Cas12a. (B) Genome-editing efficiency of whiB6 with CRISPR-Cas12a cleavage in wild-type M. marinum and its derivatives. The whiB6 crRNA plasmid (pYC1103) was transformed into the wild type and its derivatives. Transformation efficiency was defined as the total number of CFU generated per transformation. Editing efficiency was calculated as the ratio of the number of edited events to the total number of colonies analyzed by PCR and sequencing. Survival was calculated by comparing the number of transformants to the number of control transformants carrying the empty vector. (A, B) Bars represent mean values ± standard deviations from two independent experiments. (C) Deletion length distribution of the indicated genes resulting from CRISPR-Cas12-assisted genome editing. Bars represent the median deletion size for each strain.

FIG 3

NrgA is involved in CRISPR-NHEJ genome editing in mycobacteria. (A) Genome-editing efficiency in M. marinum derivatives. The whiB6 crRNA plasmid (pYC1103) was transformed into the wild-type, nrgA mutant, or nrgA-complemented strain, and 24 colonies in each group were picked for PCR and sequencing analysis. Editing efficiency was calculated as the ratio of the number of edited events to the total number of colonies tested by PCR. Normalized editing efficiency was calculated as the editing efficiency × (total CFU obtained with whiB6 targeting sgRNA/total CFU obtained with control sgRNA). (B) Deletion length distributions of whiB6 gene in wild-type and nrgA mutant strains in one experiment. Bars represent median deletion size for each strain. NS, not significant. (C, D) NrgA increases the CRISPR-NHEJ genome-editing efficiency in M. smegmatis (C) and M. tuberculosis (D). crRNA plasmids were electroporated into M. smegmatis and M. tuberculosis cells with plasmids expressing the complete NHEJ machinery (pNHEJ-Cas12a-recX for M. smegmatis and pNHEJ-recX for M. tuberculosis; Ku-NrgA-LigD) or the NHEJ machinery without NrgA (pYC1376 for M. smegmatis and pYC1654 for M. tuberculosis; Ku-LigD). For M. smegmatis, a chromosomally integrated gfp reporter gene was edited. Normalized efficiency was calculated as the frequency of GFP-negative (white) transformants × (total CFU obtained with gfp-targeting sgRNA/total CFU obtained with control sgRNA). For M. tuberculosis, 24 colonies in each group were picked for PCR and sequencing analysis. Normalized editing efficiency was calculated as the editing efficiency × (total CFU obtained with target sgRNA/total CFU obtained with control sgRNA). (A, C, D) Bars represent mean values ± standard deviations from three independent experiments. P values were determined via Student’s unpaired t test.

FIG 4

CRISPR-NHEJ-assisted genome editing in M. smegmatis. (A) CRISPR-Cas12a–NHEJ genome-editing efficiency in M. smegmatis derivatives. The gfp crRNA plasmid was transformed into wild-type and recA mutant strains harboring various plasmids. (B) RecX or RecA_mu expression mimicked recA deletion in M. smegmatis. Transformation and gfp-editing efficiency resulted from electroporation of the indicated gfp crRNA plasmids into M. smegmatis expressing the MmNHEJ machinery, Cas12a, and RecX (or RecA_mu). (C) CRISPR-Cas9_Sth1-assisted NHEJ genome editing in M. smegmatis. The gfp sgRNA plasmid was transformed into recA mutant and wild-type strains containing the indicated plasmids. Bars represent mean values ± standard deviations from three independent experiments. (D) Comparison of the Cas9_Sth1- and Cas12a-induced deletion sizes in PAM-proximal and PAM-distal regions. Transformation efficiency was defined as the total number of CFU obtained per transformation, and editing efficiency was calculated by determining the proportion of GFP-negative colonies.

FIG 5

CRISPR-NHEJ-assisted genome editing in M. tuberculosis. (A) CRISPR-Cas9_Sth1 cleavage combined with NHEJ repair leads to efficient genome editing in M. tuberculosis H37Ra. sgRNA-expressing plasmids were electroporated into M. tuberculosis cells harboring various NHEJ-expressing plasmids. Transformation efficiency was defined as the total number of CFU obtained per transformation, and editing efficiency was calculated as the ratio of the number of edited events to the total number of colonies tested. Bars represent mean values ± standard deviations from two independent experiments. (B to D) Analysis of NHEJ efficiencies induced by paired Cas9_Sth1-sgRNAs in four different orientations. Four different Cas9_Sth1-sgRNA orientations were guided by paired PAMs. C/C, C/W, W/W, and W/C orientations were defined by the positioning of the paired PAMs on either the Watson strand (W) or the Crick strand (C). (E, F) Analysis of editing efficiency induced by paired Cas9_Sth1-sgRNAs at the RD1 regions (I, II, III, and IV represent four different Cas9_Sth1-sgRNA orientations guided by paired PAMs; see Fig. S2B). (G) Simultaneous generation of double mutations in M. tuberculosis using paired sgRNAs with Cas9_Sth1 or Cas9_Sth1-ssrA. Survival was defined as the ratio of the number of CFU obtained from the indicated sgRNA to the number of CFU obtained from the control plasmids, and editing efficiency was calculated as the ratio of the number of edited events to the total number of colonies tested; at least 32 colonies were analyzed by PCR and sequenced for each group. DoDel, double deletion. (D, F) The frequencies of accurate deletion (AcDel), single deletion (SiDel), two individual deletions (TiDel), and deletion (Del) were calculated as the ratios of the number of events from each group to the total number of edited events.

FIG 6

Construction of the toxin gene knockout library. (A) Schematic overview of the construction of the CRISPR knockout toxin gene library. (B) Numbers of colonies with different sgRNAs detected. Edited colonies are shown in gray, and nonedited colonies are shown in white. (C) Genome-editing efficiency in the knockout library. Editing efficiency was calculated as the ratio of the number of edited events to the total number of colonies tested. (D) Frequencies of the indicated mutation events among the mutants detected.

FIG 7

RecA repression improves CRISPR-NHEJ genome editing in E. coli. (A) Transformation and CRISPR-NHEJ genome-editing efficiency obtained by genome editing in E. coli with or without RecX overexpression. Plasmid psgRNA-lacZ was transformed into E. coli MG1655 harboring pNHEJ and pZCas9 (or pZCas9-recX). Editing efficiency was calculated as the ratio of the number of edited events (i.e., white colonies on the X-Gal plate) to the total number of transformants. (B) Normalized editing efficiency was calculated as the ratio of edited events normalized against the transformation efficiency. Bars represent the mean values ± standard deviations from three independent experiments. For normalized editing efficiency, P values were determined by Student’s unpaired t test.