Functional Analysis of Bacteriophage Immunity through a Type I-E CRISPR-Cas System in Vibrio cholerae and Its Application in Bacteriophage Genome Engineering

Abstract

The classical and El Tor biotypes of Vibrio cholerae serogroup O1, the etiological agent of cholera, are responsible for the sixth and seventh (current) pandemics, respectively. A genomic island (GI), GI-24, previously identified in a classical biotype strain of V. cholerae, is predicted to encode clustered regularly interspaced short palindromic repeat (CRISPR)-associated proteins (Cas proteins); however, experimental evidence in support of CRISPR activity in V. cholerae has not been documented. Here, we show that CRISPR-Cas is ubiquitous in strains of the classical biotype but excluded from strains of the El Tor biotype. We also provide in silico evidence to suggest that CRISPR-Cas actively contributes to phage resistance in classical strains. We demonstrate that transfer of GI-24 to V. cholerae El Tor via natural transformation enables CRISPR-Cas-mediated resistance to bacteriophage CP-T1 under laboratory conditions. To elucidate the sequence requirements of this type I-E CRISPR-Cas system, we engineered a plasmid-based system allowing the directed targeting of a region of interest. Through screening for phage mutants that escape CRISPR-Cas-mediated resistance, we show that CRISPR targets must be accompanied by a 3' TT protospacer-adjacent motif (PAM) for efficient interference. Finally, we demonstrate that efficient editing of V. cholerae lytic phage genomes can be performed by simultaneously introducing an editing template that allows homologous recombination and escape from CRISPR-Cas targeting.

Importance: Cholera, caused by the facultative pathogen Vibrio cholerae, remains a serious public health threat. Clustered regularly interspaced short palindromic repeats and CRISPR-associated proteins (CRISPR-Cas) provide prokaryotes with sequence-specific protection from invading nucleic acids, including bacteriophages. In this work, we show that one genomic feature differentiating sixth pandemic (classical biotype) strains from seventh pandemic (El Tor biotype) strains is the presence of a CRISPR-Cas system in the classical biotype. We demonstrate that the CRISPR-Cas system from a classical biotype strain can be transferred to a V. cholerae El Tor biotype strain and that it is functional in providing resistance to phage infection. Finally, we show that this CRISPR-Cas system can be used as an efficient tool for the editing of V. cholerae lytic phage genomes.

FIG 1

Genetic context and features of the V. cholerae CRISPR-Cas system. (A) Graphic representation of CRISPR spacers in five V. cholerae classical biotype strains. Strain names are given on the left. Spacers are represented as numbered rectangles, and repeats are not shown. Numbered spacers with targets in known phages are highlighted in gray. Black rectangles represent unique spacers present in the strain indicated. Otherwise, identical spacers shared between strains are shown as rectangles with identical color schemes (combination of character color and background color), whereas different color combinations represent distinguishable spacers. Missing spacers are indicated by squares with crosses. Arrays are oriented with respect to the leader sequence (L) located on the left. (B) The CRISPR-Cas system found in classical biotype strains is part of a 16.9-kb genomic island integrated between genes equivalent to VC0289 and VC0290 in the V. cholerae El Tor strain N16961. The colored arrows indicate the genes of the genomic island, the integrase (int) and cas genes are indicated, and genes encoding hypothetical proteins are green and unlabeled for simplicity. The CRISPR locus is shown as black and gray rectangles. The location of the kanamycin resistance gene that was inserted to select for El Tor biotype transformants that acquired GI-24 via natural transformation is indicated. (C) Genetic architecture of the construct used for overexpression of CRISPR-Cas in V. cholerae El Tor. (D) The PAM sequence of the CRISPR system of V. cholerae classical biotype strains. The PAM sequence was identified after alignment of the flanking sequences of all known targets (100% match) of spacers found in the CRISPR systems of the five classical strains in panel A. No sequence conservation was detected 5′ of the target (data not shown). The sequence logo was generated using WebLogo (31).

FIG 2

CRISPR-Cas function in V. cholerae. (A) The EOP was determined for CP-T1 with each of the strains indicated compared to that with V. cholerae El Tor strain E7946. The dashed line represents the limit of detection of this assay. (B) The conjugation efficiency (number of transconjugants per donor cell) for pCP-T1 from an E. coli donor into the indicated V. cholerae strains was determined. Error bars represent standard deviations.

FIG 3

The plasmid-based CRISPR system for generating phage-resistant derivatives of V. cholerae. (A) The essential components of the pCRISPR plasmid include an IPTG-inducible CRISPR array and an MCS with several unique restriction enzyme sites for addition of editing templates. Spacers can be inserted into the CRISPR array between BsaI and/or BseRI sites using annealed oligonucleotides. (B) The oligonucleotide design for each spacer insertion site is depicted.

FIG 4

Engineered V. cholerae CRISPR confers resistance to phages. The sensitivity of each strain to a phage with the target sequence is represented as a histogram of the efficiency of plaquing, which is the ratio of the plaque count of a pCRISPR-targeting V. cholerae strain to that of the nontargeting strain. The phages used for these experiments were ICP1 (gray bars) and ICP1_2011_A (black bars). Data represent the means and standard deviations from three independent experiments. Induction of V. cholerae cas genes was achieved through an IPTG-regulated P_tac promoter upstream of the cas genes; all experiments were performed in the presence of 1 mM IPTG.

FIG 5

Mutations in the protospacer or PAM decrease CRISPR-Cas-mediated phage resistance in V. cholerae. Mutations are shown schematically beside the resulting sensitivity of each strain to the phage indicated. The EOP was determined for strain-phage combinations in which the target sequence was present (solid bars) or in which the target sequence was mutated (striped bars). Note that the PAM is immediately proximal to the protospacer; a space is present for clarity. Error bars indicate standard deviations.

FIG 6

Mutants that escape CRISPR interference arise through point mutations and deletions. (A) Chromatograms of DNA sequences of ICP1 escape mutants that formed plaques on a strain with a spacer directed against orf10 demonstrate that 5/7 escape mutants had point mutations (highlighted) and 2/7 harbored deletions that removed the CRISPR target (indicated by a blue rectangle). WT, wild type. (B) PCR analysis of the CRISPR locus of ICP1_2011_A ΔS9 mutants on a strain with a targeting plasmid directed against the phage CRISPR spacer 2 (highlighted in black). The sequenced CRISPR arrays of all 11 escape phage are schematized below the PCR analysis gel.

FIG 7

The V. cholerae CRISPR-Cas system can be used to engineer lytic V. cholerae phages. (A) Concept of phage genome editing using the CRISPR-Cas system. The CRISPR-targeting construct is cointroduced with an editing template that can recombine with the phage target sequence to permit phage escape and, thus, plaque formation. (B) The EOP was determined for strain-phage combinations in which targeting occurs in the absence (black bars) or presence (open bars) of an editing template. The lengths of the up and down arms of homology used in the editing template are indicated below the green and yellow bars, respectively. Chromatograms for representative engineered phages with deletions or insertion sizes are shown below the wild-type sequence for each editing construct. Lowercase letters highlighted in gray represent the wild-type sequence that was deleted by the editing construct.