Chromosomal Targeting by the Type III-A CRISPR-Cas System Can Reshape Genomes in Staphylococcus aureus

Figures

• 

  FIG 1. Identification of the functional CRISPR promoter region. (A) Relative transcription level of the native CRISPR array under the control of the truncated leader in the CRISPR knockout strain. The lengths of truncated leader were 404, 252, and 158 bp. Values that are significantly different from the value for the leader158 (P < 0.001) are indicated by three asterisks. (B) Artificial mini-CRISPR arrays with truncated leaders of 158 and 252 bp were constructed and transformed into the WT and cas6 knockout strains. At least three independent transformation experiments were performed, and representative plates are shown. (C) The sequences of the truncated 252-bp and 158-bp leaders, the predicted −35 and −10 promoter regions (blue), and the transcription start site (TSS) (in orange) relative to the first CRISPR repeat (red) are shown.

• 

  FIG 2. One chromosome-targeting spacer is sufficient for CRISPR targeting. (A) Schematic of two sequence regions selected as an artificial CRISPR array-targeting site. Sequences of the coding strand from 1544 to 1578 nt and from 399 to 433 nt relative to the start codon (ATG) of mecA constituted protospacer 1 and protospacer 2, respectively. (B) mecA-targeting constructs pLI-1 (one spacer), pLI-11 (two identical spacers), and pLI-12 (two different spacers) displayed similar toxicity. The transformation efficiency of the empty plasmid pLI50 (no spacer) was set at 100%. Transformations were performed three times, and average relative transformation efficiencies plus standard deviations (error bars) are shown in the graph.

• 

  FIG 3. An artificial CRISPR plasmid with spacers targeting mecA displays Cas-dependent toxicity. (A) Schematic of sequence regions selected as artificial CRISPR plasmid targeting sites. The sequence of protospacer-C is in the coding strand of mecA, whereas protospacer-T is the complementary sequence of protospacer-C in the template strand. (B) Transformation plates of the WT strain and the Δcas6 mutant strain after growth for 36 h on TSB containing chloromycetin (Chl). CRISPR plasmids contained spacers targeting the coding strand and template strand of mecA.

• 

  FIG 4. Transformants evade CRISPR targeting by different mutations. (A) Summary of different mutation types and corresponding proportions of 128 surviving clones. (B) PCR amplification for identification of large fragment deletions across SCCmec. Deletions occurred between CDS7 and CDS23. The weaker PCR bands reflected gene breaking regions. 1-12, lanes 1 to 12. (C) Schematic of two representative transformants contained about 16-kb deletion within SCCmec. The deletion regions were between CDS7 and CDS21 and between ccrC1 and ccrC2. (D) Schematic of transformants avoiding CRISPR attack by removal of the spacer repeat unit. (E) Distribution of mutations within different cas genes. Red asterisks indicate the mutation sites of single-nucleotide insertions, deletions, or substitutions.

• 

  FIG 5. Determination of the intermediate products and mature crRNAs generated from mecA-targeting CRISPR arrays by Northern blotting. (A) Size distribution of 39 spacers from six CRISPR-positive S. aureus strains. (B) Schematic of sequences with mecA-targeting spacers of different lengths (red). (C) The processed crRNAs generated from spacers of different lengths showed similar sizes. The arrows indicate the positions of mature crRNAs with sizes of ~37 nt and ~43 nt. The higher bands indicate the intermediate products. Portions of the gel were taken from three different gels and joined together. The broken lines show the spliced portions.

• 

  FIG 6. The spacer length has influence on targeting activity. (A) Schematic of base pairing between mature crRNA generated from pLI-S36 and its target sequence. Nucleotides from the spacer are highlighted in red. (B) The relative transformation efficiencies of S. aureus strain AH1 with artificial CRISPR plasmids containing spacers of different lengths. Transformations were performed at least three times. The transformation efficiency of the control (pLI50) was set at 100%. (C) PCR amplification for the detection of the mecA target gene. PCR performed for cas6 was shown as a control. (D) Relative activity of mecA-targeting constructs containing spacers of different lengths. The targeting activity of empty plasmid pLI50 was set at zero. Five independent transformants were analyzed for each construct. The values are means ± standard deviations (error bars). Values that are significantly different (P < 0.001) from the value for pLI50 are indicated by three asterisks. Values that are not significantly different (NS) from the value for pLI50 are also indicated.

• 

  FIG 7. Mutations in the crRNA 5′ tag eliminate CRISPR attack. (A) Schematic of the complementarity between the flanking sequences (positions −1 to −8) of crRNAs (bottom) and target DNA (top). The mutated nucleotides are shown in red. (B) Effects of progressive sequence mutations in the 5′-tag sequence on transformation efficiency. The transformation efficiency of the empty plasmid pLI50 was set at 100%. The construct pLI-S36, which contains the native repeat, was used as a positive control. The mutated nucleotides are shown in red. (C) The relative targeting activity of mecA-targeting constructs contained a series of mutations in the 5′-tag sequence. The targeting activity of empty plasmid pLI50 was set at zero. The construct pLI-S36 was taken as a positive control. Five independent transformants were analyzed for each construct with bars indicating standard deviations.