Degradation of Phage Transcripts by CRISPR-Associated RNases Enables Type III CRISPR-Cas Immunity

Abstract

Type III-A CRISPR-Cas systems defend prokaryotes against viral infection using CRISPR RNA (crRNA)-guided nucleases that perform co-transcriptional cleavage of the viral target DNA and its transcripts. Whereas DNA cleavage is essential for immunity, the function of RNA targeting is unknown. Here, we show that transcription-dependent targeting results in a sharp increase of viral genomes in the host cell when the target is located in a late-expressed phage gene. In this targeting condition, mutations in the active sites of the type III-A RNases Csm3 and Csm6 lead to the accumulation of the target phage mRNA and abrogate immunity. Csm6 is also required to provide defense in the presence of mutated phage targets, when DNA cleavage efficiency is reduced. Our results show that the degradation of phage transcripts by CRISPR-associated RNases ensures robust immunity in situations that lead to a slow clearance of the target DNA.

Figure 1. Csm6 is an RNase not involved in type III-A DNA degradation

(A) S. epidermidis RP62a carries a type III-A CRISPR-Cas locus that harbors four repeats (black boxes), three spacers (colored boxes) and nine cas/csm genes. cas10 and csm2–5 (in blue) encode for the Cas10-Csm ribonucleoprotein complex which has crRNA-guided DNA and RNA cleavage activities. The function of csm6 (in red) is unknown. (B) Purified Csm6 and dCsm6 were incubated with a radiolabeled ssRNA substrate. The reaction proceeded for 1 hour and aliquots were taken at 0, 15, 30, 45 and 60 minutes for PAGE and phosphorimager visualization. (C) A 5′-radiolabeled primer is used to initiate reverse transcription of the target transcript, generating a 171 nt extension product in the absence of RNA cleavage. The target is located in the pTarget plasmid under the control of a tetracycline-inducible promoter; this plasmid was introduced in different strains carrying the wild-type, Δcsm6 or Δspc (non-targeting control) CRISPR-Cas systems. Total RNA for primer extension was extracted at different times after addition of the aTc transcription inducer. Primer extension products were separated by PAGE and detected by phosphorimaging; the products indicating transcript cleavage are marked with an arrowhead. (D) pTarget plasmid DNA was extracted from cells before and after 10 hours of treatment with aTc, testing the different CRISPR-Cas backgrounds described in panel (C). See also Fig. S1.

Figure 2. Csm3 and Csm6 are required for the degradation of phage transcripts

(A) Schematic diagram of the ΦNM1γ6 genome in its linear (prophage) form, and the position of type III-A CRISPR-Cas targets used in this study. Supplemental Table S1 contains the full sequence of each target. The green arrow represents the rightwards promoter driving transcription of the lytic genes and defines the early expressed genes as those immediately downstream of this promoter. During the lytic cycle the genome is circular. The opposed arrows indicate the primers used for RT-qPCR experiments in panels B and C. (B) RT-qPCR performed on the ΦNM1γ6 gp14 transcript using total RNA collected at different times post-infection from cells carrying different type III-A CRISPR-Cas systems targeting the gp14 gene. Values for the rho gene were used for normalization. The normalized value for the measurement at 15 minutes in wild-type cells was set to 1 to obtain the relative abundance of the gp14 transcript for the rest of the data points (mean ± S.D. of four replicas). (C) Same as panel (B), but using CRISPR-Cas systems targeting the ΦNM1γ6 gp43 gene and measuring relative abundance of the gp43 transcript. (D) RNA-seq reads (Reads Per 500 bases of transcript per Million mapped reads, RPM) for transcripts in the vicinity of the gp14 target at 15 and 45 minutes post infection of cells harboring different mutations in the type III-A CRISPR-Cas system. Vertical blue line indicates target position. (E) Same as (D) but showing transcription levels in the gp43 target region. See also Fig. S2.

Figure 3. Co-transcriptional type III CRISPR-Cas targeting leads to the accumulation of phage DNA

(A) Location of the EcoRI and PsiI restriction sites used to detect phage DNA via southern blot in panel (B). The green line indicates the location of the dsDNA probe used in this assay. The gp43 target and the primers (opposed arrows) used for qPCR in panel (D) are also shown. (B) Southern blot on total DNA extracted from cells treated with ΦNM1γ6 at different times after infection and digested with EcoRI and PsiI. Cells harboring type II-A or type III-A CRISPR-Cas systems programmed to target the gp43 gene, or without CRISPR-Cas immunity were infected. The intensity, relative to type II-A targeting, of the 3.9 kb phage fragment detected is reported. (C) Ethidium bromide gel used for southern blot shown in (B). (D) qPCR performed on the ΦNM1γ6 gp43 gene using total DNA collected at different times post-infection from cells carrying different CRISPR-Cas systems. Values for the rho gene were used for normalization. The normalized value for the measurement at 15 minutes in cells harboring a type II system was set to 1 to obtain the relative abundance of the gp43 transcript for the rest of the data points (mean ± S.D. of four replicas). The R9 time point indicates that cells were refreshed with new culture broth at 9 hours post-infection and were grown for an additional 9 hours before collection of DNA for qPCR. See also Fig. S3.

Figure 4. Degradation of phage transcripts by Csm3 and Csm6 enables type III CRISPR-Cas immunity targeting late-expressed genes

(A) Staphylococci harboring different type III-A CRISPR-Cas systems targeting the gp14 gene were grown in liquid media and infected with ϕNM1γ6 phage (at 0 hours) with a multiplicity of infection of 5 viruses per bacteria. Optical density at 600 nm (OD₆₀₀) was measured for the following 12 hours to monitor cell survival due to CRISPR immunity against the phage. Representative growth curves of at least three independent assays are shown. (B) Same as panel (A) but with the CRISPR-Cas systems programmed to target gp43. (C) The different infections performed in panel (A) were plated to enumerate plaque forming units (pfu) and calculate the average burst size. Mean ± S.D. of three replicas are reported. (D) Same as panel (C) but with the CRISPR-Cas systems programmed to target gp43. (E) Survival of cells (determined by measuring growth at OD₆₀₀) carrying dcsm6/dcsm3 type III-A CRISPR-Cas systems targeting the different ϕNM1γ6 genes shown in Fig. 2A. Representative growth curves of at least three independent assays are shown.

Figure 5. Csm6 enables complete phage clearance during immunity against late-expressed genes

(A) Cells harboring Δcsm6/dcsm3 type III-A CRISPR-Cas systems targeting gp14 or gp43 were complemented with the pCsm6 plasmid, which carries the csm6 gene under the control of a tetracycline-inducible promoter. Each strain was infected with ϕNM1γ6 in the presence or absence of the aTc (0.008 μg/ml); i.e. induction of Csm6 expression. Bacterial growth was monitored by measuring for OD₆₀₀ for 10 hours. (B) The cells grown in the presence of aTc were collected, washed to remove the inducer and eliminate further expression of Csm6, and diluted (1:333) in fresh media without phage nor aTc. As a control an aliquot of the washed cells were re-inoculated in fresh media with aTc (0.008 μg/ml). Bacterial growth was monitored by measuring for OD₆₀₀ for 12 hours.

Figure 6. Csm6 is required to provide immunity against viruses with target mutations

(A) Introduction of mutations (in red) in the spacer targeting the gp43 phage gene that generate 3, 4 or 5 mismatches in the crRNA:target region. (B) Staphylococci harboring a wild-type III-A CRISPR-Cas system targeting the gp43 gene in the presence of different crRNA:target mismatches were grown in liquid media and infected with ϕNM1γ6 phage (at 0 hours) with a multiplicity of infection of 5 viruses per bacteria. Optical density at 600 nm (OD₆₀₀) was measured for the following 12 hours to monitor cell survival. Representative growth curves of at least three independent assays are shown. (C) Same as panel (B) but with cells harboring a CRISPR-Cas locus without csm6. (D) qPCR performed on the ΦNM1γ6 gp43 gene using total DNA collected at different times post-infection from cells carrying CRISPR-Cas systems targeting in the presence (4 mm) or absence (0 mm) of crRNA:target mismatches. Values for the rho gene were used for normalization to obtain the relative abundance of the gp43 gene for each data point (mean ± S.D. of four replicas). (E) pTarget plasmid DNA, harboring the gp43 target under the control of a tetracycline-inducible promoter, was extracted from cells harboring a type III-A CRISPR-Cas system without a spacer (Δspc) or with a gp43-targeting spacer with or without mismatches (4 mm or 0 mm, respectively), at different times after treatment with aTc. Plasmid DNA was visualized by agarose gel electrophoresis followed by ethidium bromide staining. (F) The gp32 spacer has a complete match in the ϕNM1γ6 genome but presents four mismatches in the ϕNM4γ4 phage. (G) Staphylococci harboring a wild-type, Δcsm6 or Δspc type III-A CRISPR-Cas system with the gp32 spacer were grown in liquid media and infected with ϕNM1γ6 phage (at 0 hours) with a multiplicity of infection of 5 viruses per bacteria. Optical density at 600 nm (OD₆₀₀) was measured for the following 12 hours to monitor cell survival due to CRISPR immunity against the phage. Representative growth curves of at least three independent assays are shown. (H) Same as (G) but following infection with phage ϕNM4γ4. See also Fig. S4.

Figure 7. A model for the requirement for transcript degradation during type III CRISPR-Cas immunity

The type III-A Cas10-Csm complex performs co-transcriptional cleavage of the target DNA and its transcripts. Within this complex, Cas10 contains the DNase activity and Csm3 is an RNase. Csm6 is another type III-A RNase that degrades target transcripts. This molecular mechanism of immunity allows for the rapid attack of the viral genome when early-expressed targets are specified by the crRNA guide, which leads to fast and efficient degradation of the invader’s genetic material and the clearance of the infection without the need of RNase activity. In contrast, the targeting of a late-expressed gene allows viral replication and transcription before DNA cleavage can occur. The accumulated genomes are not cleared efficiently by the endonuclease activity of the Cas10 complex and the degradation of phage transcripts by Csm3 or Csm6 is required to prevent the completion of the infectious cycle and the lysis of the host cell. Similarly, the presence of crRNA:target mismatches within the phage population prevents efficient DNA cleavage that also leads to the accumulation of phage genomes in the cell. In this scenario the Csm6 RNase is required for transcript degradation and survival.