Lactococcus lactis type III-A CRISPR-Cas system cleaves bacteriophage RNA

Abstract

CRISPR-Cas defends microbial cells against invading nucleic acids including viral genomes. Recent studies have shown that type III-A CRISPR-Cas systems target both RNA and DNA in a transcription-dependent manner. We previously found a type III-A system on a conjugative plasmid in Lactococcus lactis which provided resistance against virulent phages of the Siphoviridae family. Its naturally occurring spacers are oriented to generate crRNAs complementary to target phage mRNA, suggesting transcription-dependent targeting. Here, we show that only constructs whose spacers produce crRNAs complementary to the phage mRNA confer phage resistance in L. lactis. In vivo nucleic acid cleavage assays showed that cleavage of phage dsDNA genome was not detected within phage-infected L. lactis cells. On the other hand, Northern blots indicated that the lactococcal CRISPR-Cas cleaves phage mRNA in vivo. These results cannot exclude that single-stranded phage DNA is not being targeted, but phage DNA replication has been shown to be impaired.

Keywords: CRISPR; Phages; resistance.

Figure 1.

Lactococcal CRISPR-Cas synthetic spacer interference. (A) Type III-A CRISPR-Cas operon from pKLM. CRISPR-associated (cas) genes are represented as grey arrows oriented towards the direction of transcription. The leader (L) and trailer (T) sequences are shown as boxes with the repeat-spacer (CRISPR) array depicted as black diamonds.(B) Expanded view of the repeat-spacer array and flanking regions. Repeats are shown as black diamonds and spacers as rectangles. Plasmids pKLM and pG6 have a complete set of cas genes. The pKLM array has the full complement of spacers whereas pG6 array lacks spacers S2 through S9. Spacers S2, S3, and S4 have nucleotide identity and provide resistance to phages P008 and bIL170. To construct random spacer delivery vector pRS, the region demarcated by dashed lines was cloned into the broad host range vector pGhost9 and contains the 3ʹ end of cas1, lch, leader, and trailer sequence flanking a single repeat-spacer-repeat unit.

Figure 2.

The lactococcal CRISPR-Cas targets phage RNA; orf53. The middle section depicts the relevant segment of the virulent lactococcal P008 genome analyzed for the cleavage assays. In the upper panel, arrows and Northern blot show the corresponding 1.1-kb phage mRNA transcript and subsequent fragments (< 0.06-kb and 1.0-kb) that would be generated if the mRNA is targeted. In the bottom panel, arrows and Southern blot show the phage genomic DNA 4.1-kb fragment generated by SpeI-PvuII digestion and expected fragments of 2.7-kb and 1.4-kb if DNA is targeted. Equivalent amounts of DNA digest was loaded for each lane (see Materials & Methods and Supplementary Figure 1). The diminished signal intensity for the R53n, where there is phage inhibition, is a function of the amount of P008 phage DNA relative to total DNA in the extraction and digestion. The gray bars indicate the position of the probes used to detect the respective fragments. The bacterial strains used for the phage infection were L. lactis 1403S (phage-sensitive control), L. lactis 1403S-RS53c, and L. lactis 1403S-RS53n all infected or not by the strictly lytic phage P008. NI = not infected. The time in minutes post phage infection is indicated on each autoradiogram.

Figure 3.

The lactococcal CRISPR-Cas targets phage RNA; orf36. The middle section depicts the relevant segment of the virulent lactococcal P008 genome for the cleavage assays. In the upper panel, arrows and Northern blot show the corresponding 1.9-kb phage mRNA transcript and subsequent fragments (0.2-kb and 1.7-kb) that would be generated if mRNA is targeted. In the bottom panel, arrows and Southern blot show the phage genomic DNA 3.2-kb fragment generated by SpeI digestion and expected fragments of 1.62-kb and 1.59-kb if DNA is targeted. Equivalent amounts of DNA digest was loaded for each lane (see Materials & Methods and Supplementary Figure 1). The diminished signal intensity for the R36n, where there is phage inhibition, is a function of the amount of P008 phage DNA relative to total DNA in the extraction and digestion. The gray bars indicate the position of the probes used to detect the respective fragments. The strains used for the infection were L. lactis 1403S (phage-sensitive control), L. lactis 1403S-RS36c, and L. lactis 1403S-RS36n all infected or not by the strictly lytic phage P008. NI = not infected. The time in minutes post phage infection is indicated on each autoradiogram.

Figure 4.

The lactococcal CRISPR-Cas interfere with phage DNA replication; orf36 and orf53. Migration of the total DNA extractions, obtained from phage P008-infected and non-infected L. lactis strains as well as phage P008 only, digested with SpeI and PvuII restriction enzymes. The respective strains used for the infection were L. lactis 1403S (control), L. lactis 1403S-RS36c, L. lactis 1403S-RS36n, L. lactis 1403S-RS53c, and L. lactis 1403S-RS53n. The prominent banding pattern in phage infected lanes represents replication of phage DNA over host genomic DNA as a consequence of phage infection. The 5-kb band is the pRS vector which is at higher copy than host genomic DNA and contains a single SpeI site. The time 20 minutes post phage infection was used and is indicated when applicable. NI = not infected.