ABSTRACT
Many prokaryotes possess an adaptive immune system encoded by clustered regularly interspaced short palindromic repeats (CRISPRs). CRISPR loci produce small guide RNAs (crRNAs) that, in conjunction with flanking CRISPR-associated (cas) genes, combat viruses and block plasmid transfer by an antisense targeting mechanism. CRISPR-Cas systems have been classified into three types (I to III) that employ distinct mechanisms of crRNA biogenesis and targeting. The type III-A system in Staphylococcus epidermidis RP62a blocks the transfer of staphylococcal conjugative plasmids and harbors nine cas-csm genes. Previous biochemical analysis indicated that Cas10, Csm2, Csm3, Csm4, and Csm5 form a crRNA-containing ribonucleoprotein complex; however, the roles of these genes toward antiplasmid targeting remain unknown. Here, we determined the cas-csm genes that are required for antiplasmid immunity and used genetic and biochemical analyses to investigate the functions of predicted motifs and domains within these genes. We found that many mutations affected immunity by impacting the formation of the Cas10-Csm complex or crRNA biogenesis. Surprisingly, mutations in the predicted nuclease domains of the members of the Cas10-Csm complex had no detectable effect on antiplasmid immunity or crRNA biogenesis. In contrast, the deletion of csm6 and mutations in the cas10 Palm polymerase domain prevented CRISPR immunity without affecting either complex formation or crRNA production, suggesting their involvement in target destruction. By delineating the genetic requirements of this system, our findings further contribute to the mechanistic understanding of type III CRISPR-Cas systems.
INTRODUCTION
Widespread in bacteria and archaea, clustered regularly interspaced short palindromic repeats (CRISPRs) and CRISPR-associated (cas) genes constitute an adaptive immune system that uses small RNAs (crRNAs) to protect against virus infection and plasmid transfer (1, 2). A CRISPR locus consists of an array of short repeat sequences (24 to 48 nucleotides [nt] in length) that alternate with similarly sized invader-derived sequences called spacers (3). CRISPR loci are often flanked by an operon of cas genes, which encode the machinery for CRISPR immunity. There are three distinct phases of this immune system (4). During the first phase, known as adaptation, viral or plasmid challenge stimulates the addition of spacers to the archive of invader-derived sequences in the CRISPR locus (5–7). In the second phase, known as crRNA biogenesis, the repeat-spacer array is transcribed into a long precursor, which is subsequently chopped to generate mature crRNAs that consist of a single spacer flanked by partial repeat sequences on one or both sides (8–11). In the third phase, known as targeting, the crRNAs assemble with Cas proteins to form a surveillance complex that recognizes and destroys foreign genetic elements antisense to the crRNAs (12–15). Fairly prevalent in the prokaryotic world, CRISPR-Cas systems have been found in over 40% of bacteria and nearly all archaea (16, 17). However, while they serve a common purpose, CRISPR-Cas systems exhibit remarkable functional and mechanistic diversity (17). To date, more than 12 distinct CRISPR-Cas systems have been documented, and a recent comparative analysis has led to their classification into four types (I to III and U) based upon their architecture, cas gene content, and mechanisms of action (18).
Staphylococcus epidermidis RP62a is an opportunistic pathogen that harbors a type III-A CRISPR-Cas system (Fig. 1A). This system contains nine cas-csm genes that can direct antiplasmid immunity against staphylococcal conjugative plasmids (2, 19). In this system, crRNAs are generated from the precursor transcript in two steps: primary processing and maturation (10) (Fig. 1B). During primary processing, endoribonucleolytic cleavage events within repeat sequences liberate intermediate crRNAs that are 71 nt in length. During maturation, these intermediates are trimmed on the 3′ end in 6-nucleotide increments to form a collection of mature crRNAs that range from 31 to 67 nt in length (20). Of the nine cas-csm genes, csm2, csm3, and csm5 are required for maturation (10), and their gene products form a stable ribonucleoprotein complex, along with Cas10, Csm4, and mature crRNAs (20). Within this Cas10-Csm complex, Csm3 acts as a 6-nt ruler that measures the extent of maturation on the 3′ end of the crRNA intermediate.
Assembly of the Cas10-Csm ribonucleoprotein complex is required for antiplasmid CRISPR immunity. (A) Organization of the type III-A CRISPR system in S. epidermidis RP62a. This system contains nine CRISPR-associated (cas and csm) genes, four direct repeats (DR), and three spacers, the first of which (spc1) targets the nickase gene in pG0400, a staphylococcal conjugative plasmid. (B) crRNA biogenesis in S. epidermidis. Transcription of the repeat-spacer array generates a precursor crRNA that is subjected to two cleavage events: primary processing, which entails endoribonucleolytic cleavages within repeats to yield ∼71-nucleotide crRNA intermediates, and maturation, during which further trimming of the 3′ end generates mature crRNA species that are 31, 37, 43, 49, and 55 nucleotides in length. Base pairing between spc1 crRNA and the target region of pG0400 results in the prevention of plasmid transfer, presumably due to the destruction of the target by Cas nucleases. (C) Conjugation of pG0400 into S. epidermidis LM1680 harboring WT and mutant pcrispr plasmids. Conjugation was carried out by filter mating; the mean values for two independent experiments ± standard deviations (SD) obtained for recipients and transconjugants are shown. (D) Purification of Cas10-Csm from cells carrying wild-type and mutant pcrispr plasmids. His6 tags were placed on the N terminus of Csm2 or the C terminus of Csm4 in the case of the Δcsm2 mutant for complex purification from S. epidermidis LM1680 extracts using Ni2+ affinity chromatography. Purified proteins were resolved by SDS-PAGE. Protein identities were assigned as described in reference 20. (E) Nucleic acids were extracted from the indicated complexes, radiolabeled at the 5′ end, and resolved using denaturing PAGE.
While we have now begun to grasp the mechanism of crRNA maturation in this system, the machinery required for primary processing and antisense targeting remains unknown. Here, we study the genes and their conserved motifs required for CRISPR-mediated antiplasmid defense in S. epidermidis RP62a. We performed gene deletions and specific mutations that caused amino acid substitutions in the cas-csm operon and examined their effects on CRISPR immunity to plasmid conjugation, on the formation of the Cas10-Csm complex, and on the biogenesis of crRNAs. We found that many mutations affected immunity by impacting the formation of the Cas10-Csm complex or crRNA biogenesis. While mutations in the predicted nuclease domains of the members of the Cas10-Csm complex had no detectable effect on antiplasmid immunity or crRNA biogenesis, the deletion of csm6 and mutations in the cas10 Palm polymerase domain prevented CRISPR immunity without affecting either complex formation or crRNA production. This observation suggests that csm6 and the cas10 Palm domain are involved in target destruction. We discuss how the results of this study contribute to the mechanistic understanding of type III CRISPR-Cas systems.
MATERIALS AND METHODS
Bacterial strains and growth conditions.S. epidermidis LM1680, an S. epidermidis RP62a derivative lacking the CRISPR-Cas locus (20), and Staphylococcus aureus OS2 (21) strains were grown in brain heart infusion (BHI) medium (Difco). When required, the medium was supplemented with antibiotics as follows: neomycin (15 μg/ml) for selection of S. epidermidis LM1680, chloramphenicol (10 μg/ml) for selection of pcrispr-based plasmids, and mupirocin (5 μg/ml) for selection of pG0400. Escherichia coli BL21(DE3) Codon Plus cells (Stratagene) were grown in LB supplemented with chloramphenicol (34 μg/ml). When appropriate, the medium was also supplemented with kanamycin (50 μg/ml) to select for pET28b-His10Smt3-Cas6.
Plasmid construction.pcrispr was constructed by fusing plasmids pLM304 (10) and pC194 (22), as previously described (20). His6 tags and amino acid substitutions were introduced into pcrispr by inverse PCR using the primers listed in Table S1 in the supplemental material. Restriction cut sites were included on the primers for the generation of deletions (PspOMI and EagI), His6 tags (NheI), and amino acid substitutions (Cas10H14A,D15A, PvuI; Cas10G584A,G585A'D586A,D587A, Csm3G182A,G183A,G185A, Csm5G261A,G263A,G265A, and Cas6H41A,H42A, PstI; Csm4G194A,G196A, NdeI; Cas6G237A,G239A, BmtI) to facilitate ligation. Following inverse PCR, the products were purified using a PCR purification kit (Qiagen) and were either 5′ phosphorylated using T4 polynucleotide kinase (for blunt-end ligation) or cleaved with the appropriate restriction enzyme(s) (NEB). Restriction digests were heat-inactivated according to the manufacturer's recommendations. The digested or phosphorylated PCR products were then circularized using T4 DNA ligase (NEB). All constructs were transformed into S. aureus OS2 prior to transformation into S. epidermidis LM1680. At least two chloramphenicol-resistant transformants were selected for each construct, and plasmids were purified. To confirm the intended mutations, the entire CRISPR locus was sequenced using primers L19, W13, W14, L35, W15, W16, W17, W18, T17, and W19 (see Table S1 in the supplemental material). At least two isolates of mutant plasmids were prepared from S. aureus and transformed into S. epidermidis LM1680.
pET28b-His10Smt3-cas6 was constructed for expression and purification of S. epidermidis Cas6 in E. coli by inserting an S. epidermidis cas6 PCR product into the pET28b-His10Smt3 multiple-cloning site. Briefly, cas6 was amplified from pcrispr-cas with primers PS5 and PS6 (see Table S1 in the supplemental material), and both the PCR product and pET28b-His10Smt3 were digested with BamHI (NEB) and XhoI (NEB). The digested PCR product and linearized vector were gel purified, combined, and ligated with T4 DNA ligase (NEB) and then transformed into E. coli BL21(DE3) Codon Plus cells (Stratagene). The identity of the cloned DNA fragment was confirmed by sequencing using primers T7P and T7T (see Table S1 in the supplemental material).
Conjugation.Conjugation was carried out by filter mating as previously described (2). Confirmation of the presence of the desired plasmids in transconjugants was achieved by extracting the DNA of at least two colonies and performing PCR with suitable primers (L70/L71 to confirm pG0400 and various primers as specified in Table S1 in the supplemental material to confirm pcrispr mutations).
Cas10-Csm purification from S. epidermidis.Cas10-Csm complexes were purified from S. epidermidis LM1680 strains harboring the appropriate pcrispr construct as previously described (20).
Extraction and visualization of crRNAs from the Cas10-Csm complex.CrRNAs were extracted from 2 to 4 μg of Cas10-Csm complex by adding 750 μl TRIzol (Life Technologies) and following the manufacturer's protocol for RNA extraction. Purified RNA was end labeled using T4 polynucleotide kinase (NEB) and [γ-32P]ATP. RNAs were resolved on a 12% polyacrylamide gel and visualized by exposure to a storage phosphor screen. The screen was scanned using a Typhoon 9400 phosphorimager (GE Healthcare).
Cas6 purification from E. coli.Cultures (4 liters) of E. coli BL21(DE3) Codon Plus cells (EMD Millipore) containing the pET28b-His10Smt3-Cas6 plasmid were grown at 37°C in Luria-Bertani medium containing 50 μg/ml kanamycin and 34 μg/ml chloramphenicol until the optical density at 600 nm (OD600) reached 0.6. The cultures were adjusted to 0.3 mM isopropyl-1-thio-β-d-galactopyranoside, and incubation was continued for 16 h at 17°C with constant shaking. The cells were harvested by centrifugation, and the pellets were stored at −80°C. All subsequent steps were performed at 4°C. Thawed bacteria were resuspended in 30 ml of buffer A (50 mM Tris-HCl, pH 7.5, 1.25 M NaCl, 200 mM Li2SO4, 10% sucrose, 15 mM imidazole) containing one Complete EDTA-free protease inhibitor tablet (Roche). Triton X-100 and lysozyme were added to final concentrations of 0.1% and 0.1 mg/ml, respectively. After 30 min, the lysate was sonicated to reduce viscosity. Insoluble material was removed by centrifugation for 30 min at 15,000 rpm in a Beckman JA-3050 rotor. The soluble extract was mixed for 1 h with 2 ml of Ni2+-nitrilotriacetic acid-agarose resin (Qiagen) that had been preequilibrated with buffer A. The resin was recovered by centrifugation and then washed with 40 ml of buffer A, followed by washing with 5 ml of 3 M KCl solution. The resin was subsequently resuspended in 25 ml of buffer A and then poured into a column. The column was eluted stepwise with 3-ml aliquots of IMAC buffer (50 mM Tris-HCl, pH 7.5, 250 mM NaCl, 10% glycerol) containing 50, 100, 200, 350, and 500 mM imidazole. The 350 mM imidazole eluates containing the His-tagged Cas6 polypeptide were pooled, aliquoted, and frozen. Protein concentrations were determined by using the Bio-Rad dye reagent with bovine serum albumin (BSA) as the standard.
Cas6 in vitro nuclease activity assay.Cas6 nuclease activity was tested against a 185-nucleotide RNA substrate containing the first 86 nucleotides of the CRISPR leader, the first direct repeat, spacer 1, and 28 nucleotides of the second repeat (5′-ATAATCTTGTACTAGTGATTGTCATATTTTTTGACAGCAAAAATGATGCTTGAAATATAGTTGTGATGGCATTTGTTAAAGTATCGGATCGATACCCACCCCGAAGAAAAGGGGACGAGAACACGTATGCCGAAGTATATAAATCATCAGTACAAAGGATCGATACCCACCCCGAAGAAAAGGGG-3′). This substrate was generated and internally labeled using the HiScribe T7 in vitro transcription kit (NEB), [α-32P]ATP, and a template created by PCR amplification of a portion of the pcrispr repeat-spacer array (using primers A130 and A131 [see Table S1 in the supplemental material]). The radiolabeled substrate was PAGE purified prior to use. Nuclease assays were performed by combining the indicated amounts of Ni2+ affinity-purified Smt3-Cas6 and PAGE-purified radiolabeled substrate in nuclease buffer (10 mM Tris-acetate, pH 8.0, 14 mM Mg acetate, 60 mM K acetate, and 1 mM dithiothreitol [DTT]) and incubating at 37° for 20 min.
RESULTS
Assembly of the Cas10-Csm ribonucleoprotein complex is required for antiplasmid CRISPR immunity.S. epidermidis RP62a harbors a type III-A CRISPR-Cas system (Fig. 1A) that can prevent the horizontal transfer of the staphylococcal conjugative plasmid pG0400 in a spacer 1 (spc1)-dependent manner (2). In addition to cas1 and cas2, genes known to be involved in the acquisition of new spacers (5, 7), this system harbors seven additional cas-csm genes whose involvement in CRISPR targeting is unclear. To test each gene for its contribution to targeting, we used pcrispr, a plasmid that contains the entire S. epidermidis RP62a CRISPR-Cas system cloned into the multicopy staphylococcal plasmid pC194 (22), and created in-frame deletions of each cas-csm gene. We introduced these plasmids into S. epidermidis LM1680, an RP62a derivative that lacks the CRISPR-Cas locus region (20). We then tested each in-frame deletion plasmid for its ability to prevent the transfer of the conjugative plasmid pG0400, which contains a spc1 target in its nickase gene (2) (Fig. 1C).
When expressed in S. epidermidis LM1680, wild-type pcrispr attenuates the conjugative transfer of pG0400 by 3 orders of magnitude (Fig. 1C and Table 1). The background level of transconjugants (∼102 to 103 CFU) seen in the presence of wild-type pcrispr, but absent when using wild-type S. epidermidis RP62a as the recipient (2), is due to plasmid rearrangements that eliminate part or all of the CRISPR-Cas locus (data not shown). As expected from their demonstrated role in spacer acquisition, deletion of cas1 and cas2 had no effect on CRISPR interference (Fig. 1C and Table 1). In contrast, deletion of cas10, csm2, csm3, csm4, csm5, csm6, or cas6 abrogated CRISPR immunity.
Conjugation efficiencies of S. epidermidis LM1680 containing pcrispr constructs with different mutations
We reported previously that Cas10, Csm2, Csm3, Csm4, and Csm5 assemble with crRNAs to form a type III-A Cas10-Csm complex (20) that mediates CRISPR immunity. We also showed that this complex cannot be formed in the absence of Csm3. We hypothesized that, similarly, the absence of the other members of the complex could affect its assembly and thus CRISPR immunity. To test this, we added His6 tags to different members of the complex in each deletion construct and used Ni2+ affinity chromatography to isolate the tagged species (Fig. 1D). Consistent with our previous result, we were unable to recover a complex in the absence of Csm3. Similarly, in the absence of Cas10 or Csm4, we were only able to recover the His6-tagged species (His6-Csm2), showing that the remainder of the complex either fails to assemble or is unstable in the absence of Cas10 and Csm4. Either scenario would explain the immunity defect of cas10 and csm4 mutants. In contrast, we recovered a subcomplex containing all other members when Csm2 or Csm5 was deleted. Previously, we showed that mutant Δcsm2 and Δcsm5 cells accumulated crRNA intermediates but lacked mature crRNA species (10). Therefore, we extracted and end labeled crRNAs from these subcomplexes and showed that they carry only intermediate crRNAs (Fig. 1E). These results suggest that maturation of crRNAs is a prerequisite for CRISPR plasmid targeting.
Although not part of the Cas10-Csm complex, the deletion of csm6 and cas6 also prevented antiplasmid immunity. We previously showed that csm6 is not required for crRNA biogenesis (10), suggesting a specific role for Csm6 in the targeting phase of antiplasmid CRISPR immunity. Cas6, on the other hand, is essential for crRNA biogenesis.
Cas6 is necessary and sufficient for primary processing of the crRNA precursor in S. epidermidis.We next wanted to examine the precise role of Cas6, which is nearly universal in type I and III CRISPR-Cas systems (18). In these systems, members of the Cas6 family are considered the primary processing endonucleases (9, 11, 23). There are exceptions, however; in the type I-C system present in Bacillus halodurans, Cas5d was found to carry out primary processing (24). In S. epidermidis, we showed that crRNAs cannot be detected in Δcas6, Δcas10, or Δcsm4 mutant cells (10). Therefore, it is formally possible that the primary processing is performed by either Cas6 or the Cas10-Csm complex. To test if S. epidermidis Cas6 is indeed the primary processing endonuclease, we created a His10-SUMO fusion of Cas6 and purified it from E. coli using Ni2+ affinity chromatography. Cleavage of the SUMO moiety greatly diminished Cas6 solubility, and since this tag has been reported to be innocuous to nuclease activity (24), we used the His10-SUMO-Cas6 fusion protein for our assays (Fig. 2A). Previously, we determined the precise recognition sequence and cleavage site within repeats for the primary processing machinery (10). Accordingly, we generated an internally labeled RNA substrate that contains a single cleavage site for primary processing (Fig. 2B). The substrate consists of the leader region of the CRISPR array; a full repeat containing a primary processing site, spc1; and a partial repeat that ends at the junction of the second primary processing site. Primary processing of this substrate should result in the production of a 71-nucleotide RNA identical in sequence to the spc1 crRNA intermediate. These cleavage products were indeed observed when we incubated the radiolabeled substrate with Cas6 (Fig. 2C). We observed similar cleavage within longer substrates containing all three spacers and intervening repeats (see Fig. S1 in the supplemental material), indicating that this nuclease is necessary and sufficient for cleavage within the primary processing site and for liberation of intermediate crRNAs. Cleavage occurs in the absence of divalent metal ions (see Fig. S1 in the supplemental material), as has been previously reported for other members of the Cas6 family (8, 9, 11). To further corroborate that Cas6 functions as the primary processing endonuclease in vivo, we generated alanine mutations in the predicted catalytic histidines (H41 and H42) and in glycine residues predicted to form a glycine-rich loop (see below), a conserved region that plays a role in crRNA recognition by Cas6 (G237 and G239) (9, 25) (Fig. 3A; see the supplemental material). Similar to the phenotype of Δcas6 mutants, cas6H41A,H42A and cas6G237A,G239A displayed complete loss of crRNAs (data not shown) and loss of CRISPR immunity (Fig. 2D).
Cas6 is sufficient for primary processing within repeat sequences. (A) SDS-PAGE of purified His10-SUMO-Cas6. Cas6 was cloned into a pET28b vector and expressed and purified from E. coli BL21. (B) Diagram of the substrate used for the nuclease assay. It contains 86 nt of leader sequence followed by a full repeat (blue square; 36 nt), spc1 (yellow; 35 nt), and a partial second repeat that lacks the sequences required for primary processing (28 nt). There is a single Cas6 cleavage site in the substrate within the first repeat (arrowhead) that produces two RNAs of 114 and 71 nt. The substrate was generated and internally labeled by in vitro transcription. (C) Cleavage of the radiolabeled substrate by Cas6. Different concentrations of Cas6 (0, 5, 50, or 500 nM) were incubated with the substrate for 20 min at 37°C, and RNA was extracted and visualized by denaturing PAGE. (D) spc1-directed CRISPR interference against pG0400 in the presence of the indicated Cas6 mutations. S. epidermidis LM1680 isolates expressing WT and cas6 mutant pcrispr plasmids were used as recipients. Conjugation was carried out by filter mating; the mean values for two independent experiments ± SD obtained for recipients and transconjugants are shown.
The Csm3 and Csm5 G-rich loops are required for protein stability. (A) Diagram of Cas proteins for which catalytic activity has been predicted, including the RAMPs and the Cas10 “CRISPR polymerase.” Putative catalytic histidines and aspartates are indicated by red and black asterisks, respectively. Glycine-rich regions (G-rich loops) in the RAMPs are indicated by red rectangles. Also highlighted are the HD nuclease domain (green) and Palm polymerase domain (orange) of Cas10 and the RAMP-like RRM domain (yellow) of the RAMPs. (B) Conjugation of pG0400 S. epidermidis LM1680 cells harboring wild-type and mutant pcrispr plasmids carrying mutations in the G-rich loops of Casm3, Csm4, and Csm5. Conjugation was carried out by filter mating; the mean values for two independent experiments ± SD obtained for recipients and transconjugants are shown. (C) Purification of Cas10-Csm wild type and G-rich loop mutant complexes. His6 tags were placed on the N terminus of Csm2, and constructs were expressed in S. epidermidis LM1680. Whole-cell lysates were subjected to Ni2+ affinity chromatography, and protein extracts were resolved by SDS-PAGE. (D) CrRNAs were extracted from the indicated His-tagged Cas10-Csm complexes and visualized using denaturing PAGE. The length of each crRNA species in nucleotides is noted on the left of the gel..
The Csm3 and Csm5 G-rich loops are required for protein stability.In order to understand the molecular mechanism of action of the Cas10-Csm complex, we tested the effects of mutations in different domains of each member of the complex for which bioinformatics predictions about their functions are available (4). Along with Cas6, Csm3, Csm4, and Csm5 belong to the RAMP (repeat-associated mysterious proteins) superfamily. RAMPs are abundantly present in type I and III CRISPR-Cas systems, which may harbor one or several proteins from the family (4). These proteins contain one or more RAMP domains that possess folds similar to RNA recognition motif (RRM) domains. They also possess predicted nuclease activities and glycine (G)-rich loops, which are the signature motif of all RAMPs (4) (Fig. 3A; see the supplemental material).
We began with an assessment of the importance of G-rich loops, which have been reported to play a role in RNA recognition and protein structure (9, 25) in members of the Cas6 RAMP family. We introduced alanine substitutions in key glycine residues that form the G-rich loops of Csm3, Csm4, and Csm5 and assayed the mutant cells for antiplasmid CRISPR immunity, complex formation, and crRNA content (Fig. 3). We found that while the mutations introduced into the G-rich loop in Csm4 have no effect on CRISPR function, the G-rich loops in Csm3 and Csm5 are required for CRISPR function against plasmids (Fig. 3B and Table 1). Purification of the mutant complexes and analysis of the associated crRNAs showed that while the Csm4G194A,G196A complex is similar to the wild type, the Csm3G182A,G183A,G185A and Csm5G261A,G263A,G265A mutations mirror the phenotype of the Δcsm3 and Δcsm5 gene deletions, respectively (Fig. 1D and Fig. 3C). We previously reported that the absence of csm3 prevents the formation of the Cas10-Csm complex (20), and we could not purify any members of the complex when Csm3G182A,G183A,G185A was expressed. Similar to the csm5 deletion, complexes purified from cells expressing Csm5G261A,G263A,G265A lacked Csm5 and were associated only with intermediate crRNA, but not mature, species. These results suggest that the G-rich loops in Csm3 and Csm5 are important for the stability of the proteins, likely by maintaining their structural integrity and/or solubility.
The Cas10 Palm domain is required for antiplasmid immunity.We next wanted to explore the importance of the predicted catalytic domains for antiplasmid CRISPR immunity. In addition to the G-rich loops, the RAMPs are hypothesized to possess conserved histidines that might be important for nucleolytic activity involved in the cleavage of crRNAs during maturation or in antisense targeting (4) (Fig. 3A; see the supplemental material). We have previously shown that the predicted catalytic histidine in Csm3 (H18) is not essential for CRISPR immunity (20), but the roles of the predicted nuclease domains of Csm4 and Csm5 remain unknown. To test this, we generated alanine substitutions in Csm4 and Csm5 catalytic histidines (Csm4H17A and Csm5H17A) and tested the mutant cells for their ability to prevent plasmid conjugation (Fig. 4A). These mutations had no detectable immunity phenotype, suggesting that the histidine residues are not involved in the nucleolytic destruction of the invading plasmid and are possibly not involved in the maturation of crRNAs. To unequivocally test this, we purified complexes carrying either Csm4H17A or Csm5H17A and labeled the copurifying crRNAs to visualize them after PAGE (Fig. 4C). In both cases, the 6-nt interval pattern of mature crRNAs was indistinguishable from those extracted from wild-type complexes, demonstrating that the Csm4 and Csm5 predicted nuclease domains do not cleave intermediate crRNAs.
The Cas10 Palm domain is required for antiplasmid CRISPR immunity. (A) Conjugative transfer of pG0400 S. epidermidis LM1680 recipient cells harboring wild-type and mutant pcrispr plasmids. Conjugation was carried out by filter mating; the mean values for two independent experiments ± SD obtained for recipients and transconjugants are shown. (B) Purification of wild-type and Palm domain mutant Cas10-Csm complexes. His6 tags were placed on the N terminus of Csm3 (for the wild-type construct) or the N terminus of Csm2 (For the Cas10 Palm mutant construct). Constructs were expressed in S. epidermidis LM1680, whole-cell lysates were subjected to Ni2+ affinity chromatography, and the purified proteins were resolved by SDS-PAGE. (C) CrRNAs were extracted and visualized from each of the His-tagged Cas10-Csm complexes containing the indicated mutations. The length of each crRNA species in nucleotides is noted on the left of the gel.
In addition to the RAMPs, Cas10 harbors predicted catalytic domains. Cas10 is the signature protein of type III CRISPR-Cas systems and the largest subunit of the type III effector complexes (13, 20, 26). Cas10 possesses a predicted HD (histidine-aspartate) nuclease domain on its N terminus that might function during crRNA biogenesis or the targeting stage of antiplasmid CRISPR immunity. It also contains sequence features that resemble the Palm domain commonly found in DNA polymerases and nucleotide cyclases (20) (Fig. 3A). The Cas10 Palm domain harbors a degenerate GGDEF motif (GGDD), usually present in nucleotide cyclases, with an unknown function in the CRISPR interference pathway (27–29). We made mutants in both domains, Cas10H14A, Cas10H14A,D15A, and Cas10G548A,G585A,D586A,D587A, and tested them for antiplasmid immunity, complex formation, and crRNA biogenesis (Fig. 4). While mutations in the HD nuclease domain produced phenotypes indistinguishable from the wild type, mutation of the Palm domain prevented CRISPR immunity (Fig. 4A). This domain might play a structural role, important for Cas10 folding and/or stability, or it might play a catalytic role. In order to begin to distinguish between the two possibilities, we isolated Cas10G548A,G585A,D586A,D587A complexes, speculating that if the folding and/or stability of the Cas10 Palm domain mutant was perturbed, then we should not be able to isolate a complex, similar to the case of the Δcas10 mutant (Fig. 1E). Interestingly, the Cas10 Palm domain mutant complex remained intact (Fig. 4B), suggesting that the Cas10 Palm polymerase domain might be contributing a catalytic function, either to crRNA biogenesis or antiplasmid targeting. To distinguish between these two possibilities, we extracted and end labeled crRNAs from the complexes (Fig. 4C). The Cas10G548A,G585A,D586A,D587A mutant complex harbored wild-type crRNAs, suggesting that the Cas10 Palm domain might play a functional role downstream of crRNA biogenesis, potentially during target recognition and/or cleavage.
DISCUSSION
Here, we investigate the role in antiplasmid CRISPR immunity of the genes present in the type III-A CRISPR-Cas system of S. epidermidis RP62a and their conserved domains. There are nine cas-csm genes in the system (Fig. 1A). Among them, cas1 and cas2 have been implicated in the acquisition of new spacer sequences upon phage or plasmid infection (5–7) and therefore are not required for antiplasmid immunity in S. epidermidis. The other seven genes are essential to prevent plasmid transfer.
One of these essential genes is cas6. Cas6 family members are nearly universal among type I and III CRSIPR-Cas systems (18), and while Cas6 proteins in different systems exhibit remarkable sequence and structural diversity, they are conserved at the functional level as dedicated nucleases for crRNA processing (30). In the type III-A system in S. epidermidis, we show that Cas6 is necessary and sufficient for the first step in crRNA biogenesis: the cleavage of the precursor crRNA into intermediate crRNAs. Without Cas6 activity, the crRNA guides that mediate target recognition are not produced and the incoming conjugative plasmid is not challenged by CRISPR immunity. Essential for Cas6 activity are its nuclease and G-rich loop domains. Thus, S. epidermidis Cas6 behaves very similarly to one of the best-studied members of its family (Pyrococcus furiosus Cas6), for which previous studies have shown the importance of the catalytic histidine and G-loop glycine residues for crRNA cleavage and binding, respectively (9, 25).
Also essential for CRISPR immunity are the genes encoding Cas10, Csm2, Csm3, Csm4, and Csm5, which associate with intermediate and mature crRNAs to form the Cas10-Csm ribonucleoprotein complex (20). Csm3, Cas10, and Csm4 are required for complex formation. In the absence of Csm2 or Csm5, on the other hand, a complex containing the remaining proteins is assembled. These complexes, however, lack mature crRNAs and fail to prevent the transfer of the target plasmid. We believe these data demonstrate that maturation is directly or indirectly linked to targeting. One possibility is that maturation is required for target destruction. This hypothesis is supported by in vitro cleavage assays carried out with purified complexes of the type III-B CRISPR-Cas system of P. furiosus (13). In this archaeon, the target cleavage site was shown to be exactly 14 nt upstream from the 3′ end of the mature crRNA. It is therefore possible that the extended 3′ end of the intermediate crRNA (which is not complementary to the target) might interfere with the targeting cleavage event. Alternatively, Csm2 and Csm5 themselves could be required for target recognition and/or cleavage.
The members of the Cas10-Csm complex have been subjected to extensive bioinformatic analyses that determined the presence of conserved residues within domains with various predicted functions (4, 17, 27). Cas10, Csm3, Csm4, and Csm5 display conserved histidines that were hypothesized to participate in nucleic acid cleavage and/or degradation. Within the CRISPR immunity pathway, nuclease activity could be involved in the maturation of crRNAs and/or target destruction. We tested these possibilities by analyzing the effects of alanine mutations in the conserved histidine residues. We previously showed that the Csm3 predicted catalytic histidine did not have any detectable phenotype. In addition, it was shown that the HD nuclease domain of Cmr2, a Cas10 ortholog encoded by the type III-B CRISPR-Cas system of P. furiosus, also lacks a detectable effect on the in vitro cleavage of RNA targets (31). Our results indicate that the predicted nucleolytic histidines of Cas10, Csm4, and Csm5 do not affect antiplasmid CRISPR immunity, complex formation, or crRNA maturation. These results corroborate previous findings that suggest that an external RNase is responsible for crRNA maturation (20). The identity of the nuclease involved in plasmid target destruction, however, remains unknown.
Cas10 is also known as the “CRISPR polymerase,” because in addition to the HD nuclease domain, it contains a region with homology to the Palm domain of DNA and RNA polymerases (4, 17), with a conserved tetrapeptide motif (GGDD) commonly found in nucleotide cyclases (32). The function of this domain in CRISPR immunity is unknown. We mutated the GGDD motif and found that the mutant was unable to prevent plasmid transfer. The mutation did not affect the formation of the Cas10-Csm complex or its crRNA content, suggesting that the Cas10 Palm domain is important during targeting, i.e., DNA recognition and/or cleavage. Structural analyses of P. furiosus Cmr2, a Cas10 ortholog, revealed that the aspartates of the GGDD motif coordinate two Ca2+ ions and an ADP molecule (31), similar to the case of adenylyl cyclases (31, 33). However, neither polymerase nor cyclase activity could be reconstituted with this protein (33), and asparagine substitutions for the two conserved aspartates did not impair RNA cleavage in vitro by the Cmr complex (31). Because the type III-B Cmr CRISPR-Cas system of P. furiosus cleaves RNA molecules, as opposed to the type III-A CRISPR-Cas system of S. epidermidis, which targets DNA invaders (2, 19), it is possible that the GGDD motif is required specifically for DNA targeting. Studies on the type I-E CRISPR-Cas system of E. coli lend some support for this hypothesis. In this system, five Cas-Cse genes form a complex for antiviral defense (cascade) that targets DNA (8, 15). In this complex, CasA is the largest subunit and is responsible for DNA scanning and nonself target recognition (34). While Cas10 and CasA share no readily detectable sequence or structural homology, similarities in their overall architecture have been noted (18). We speculate that in our system, Cas10 might play a role analogous to that of CasA, with the Cas10 Palm domain being required for sliding along the DNA and scanning for targets.
In this work, we also explored the importance of the G-rich loops present in Csm3, Csm4, and Csm5. With the exception of the G-rich loop of P. furiosus Cas6, involved in RNA recognition (9, 25), the function of these conserved glycines is unknown. Here, we showed that alanine substitutions in the G-rich loop of Csm4 have no discernible phenotype. In contrast, mutations in the G-rich loops of Csm3 and Csm5 have the same effect as the gene deletion. It has been observed that the G-rich loops present in Csm3-Csm5 and Cas6 are different (4), indicating they might play a role that is distinct from the RNA recognition function of the Cas6 G-rich loop. Our observations suggest that in Csm3 and Csm5, the G-rich loops are important for protein folding/and or stability. Additional functions, impossible to detect by this type of mutational analysis, are also possible.
Finally, we found that csm6 is also essential for antiplasmid CRISPR immunity. Because neither the formation of the Cas10-Csm complex nor the biogenesis of crRNAs is compromised in cells lacking csm6, we speculate that it is involved in plasmid targeting. Several bioinformatic predictions have been made about the function of this gene. One study suggested that Csm6 acts as a transcription factor (17); however, the presence of wild-type levels of the complex proteins and crRNAs in the Δcsm6 mutant argues that Csm6 does not regulate the transcription of the CRISPR-Cas locus. More recent work showed that Csm6 orthologs harbor a novel version of the HEPN (higher eukaryotes and prokaryotes nucleotide-binding) domain, which might be involved in RNA processing or other nonspecific nuclease activity (35). Similar HEPN domains in prokaryotes are essential components of abortive infection and toxin-antitoxin systems, raising the question of whether Csm6 plays a role in cell death when a foreign invader is detected. Future genetic and biochemical analyses will explore this idea.
ACKNOWLEDGMENTS
We thank David Bikard for critical discussion of the paper.
P.S. is supported by a Helmsley Postdoctoral Fellowship for Basic and Translational Research on Disorders of the Digestive System at The Rockefeller University. L.A.M. is supported by the Searle Scholars Program, the Rita Allen Scholars Program, an Irma T. Hirschl Award, a Sinsheimer Foundation Award, and an NIH Director's New Innovator Award (1DP2AI104556-01).
We have no conflicting financial interests.
FOOTNOTES
- Received 23 September 2013.
- Accepted 26 October 2013.
- Accepted manuscript posted online 1 November 2013.
- Address correspondence to Luciano A. Marraffini, marraffini{at}rockefeller.edu.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/JB.01130-13.
REFERENCES
- Copyright © 2014, American Society for Microbiology. All Rights Reserved.
