tRNA anticodon cleavage by target-activated CRISPR-Cas13a effector

E. coli strains and plasmids

E. coli C3000 (wild-type) cells were transformed with two plasmids: CRISPR Cas13 plasmid containing the Leptotrichia shahii (Lsh) Cas13a locus and RFP plasmid for inducible expression of red fluorescent protein (RFP). The targeting cells contained the CRISPR Cas13a plasmid encoding crRNA spacer targeting the RFP mRNA. The cells contained the CRISPR Cas13a plasmid encoding crRNAs with no matching sequences in E. coli genome and the plasmids were used as the nontargeting control. CRISPR Cas13 plasmids (pC002, pC003, pC008, and pC003RFP1) described previously3 were gifts from the Feng Zhang lab. Additional CRISPR Cas13a plasmids were constructed using Golden Gate cloning on the base of pC003 plasmid containing BsaI sites for spacer cloning (Supplementary Table 1). All CRISPR Cas13 plasmids have a pACYC184 backbone carrying a chloramphenicol resistance gene. RFP plasmid (pC008) is pBR322 derivative carrying _bla gene for ampicillin resistance and rfp gene which expression is under control of tetR-promoter induced by anhydrotetracycline. E. coli Δ10 strain contained genomic deletions of 10 type II toxin-antitoxin systems (ΔchpSB, ΔmazEF, ΔrelBE, ΔyefM-yoeB, ΔdinJ-yafQ, ΔyafNO, ΔprlF-yhaV, ΔhicAB, ΔhigBA, ΔmqsRA) described previously19 was a gift from the Van Melderen lab. E. coli Δ10 strain was transformed with two plasmids: CRISPR Cas13 plasmid containing the LshCas13a locus encoding Cas13a and crRNA spacer targeting the RFP mRNA, and RFP plasmid for inducible expression of RFP to construct targeting Δ10 cells. In nontargeting Δ10 cells, CRISPR Cas13 plasmid encodes crRNA with no matching sequences in E. coli genome and the plasmids.

Cell growth assay

To investigate the effect of the RFP mRNA targeting on the cell growth rate, individual colonies of transformed targeting and nontargeting wild-type or Δ10 E. coli cells were grown overnight in LB supplemented with 25 μg/ml chloramphenicol and 100 μg/ml ampicillin. Cell cultures were diluted 1:100 in fresh LB containing chloramphenicol and ampicillin and grown for 1 hour at 37°C with continuous shaking. After 1 hour, RFP expression was induced with 500 ng/ml anhydrotetracycline and OD600 measurements were taken every 30 minutes. All growth experiments were performed at least three times.

Time-lapse microscopy

For microscopy, E. coli cells supplemented with CRISPR Cas13 plasmid and RFP plasmid were grown under the same conditions as described above – overnight cultures were diluted 1:100 and grown for 1 hour at 37°C. Aliquots of the cell cultures mixed with 100 nM YOYO-1 dye were dropped on an LB-1.5% agarose block supplemented with 25 μg/ml chloramphenicol and 100 μg/ml ampicillin, and 500 ng/ml anhydrotetracycline for induction of RFP expression. Two agarose blocks containing targeting and nontargeting cells were placed into one microscope chamber and cell growth of two cultures was simultaneous monitored under induced conditions. The experiment was done in triplicates. The Nikon Ti-E inverted microscope was equipped with Andor’s Zyla 4.2 sCMOS camera, Semrock filter Set YFP-2427B for green fluorescence detection and custom-made incubation system to maintain cells at 37°C. Image analysis was done using ImageJ software28. For each of the three replicates, the fate of at least 100 cells were monitored, and two cell types were determined - dividing cells that formed micro-colonies over the course of the experiment, and non-dividing cells that did not form colonies nor had any visible changes in cell morphology. Both cell types were represented as a percentage of the total number of cells counted for each replicate. To distinguish between live and dead cells, green-fluorescent membrane-impermeant YOYO-1 dye was used. The dye cannot penetrate live cells but can penetrate dead cells to stain the DNA, making dead cells fluoresce green. Number of dead cells were counted in targeting and nontargeting cultures and at least 100 cells were analyzed individually for three replicates.

Metabolic labelling and autoradiography

Targeting and nontargeting E. coli C3000 cells were grown in M9 minimal media containing 18 amino acids without methionine and cysteine till 0.1 OD600 at 37°C with continuous shaking. RFP expression was then induced with 500 ng/ml anhydrotetracycline. 500 μl-aliquots were taken at time 0 and then at 10, 30, 60, and 120 min post-induction and mixed with 25 μl of a solution of thymidine (10 μg/ml), or uridine (50 μg/ml), or a mixture of methionine and cysteine (10 μg/ml each) containing 1 μCi of radioactive methyl-3H-thymidine (6.7 Ci/mmol), or 5,6-3H-uridine (35-50 Ci/mmol), or L-35S-methionine (1135 Ci/mmol), (Perkin-Elmer, USA), respectively. Pulse-labelling was carried out at 37°C for 2 min with continuous shaking at 300 rpm, followed by the addition of 100 μl cold 40% TCA to stop the reaction. Samples were filtered through glass microfiber filters (GE Healthcare Whatman), and the filters were washed twice with 1 ml cold 10% TCA and 1 ml cold 100% ethanol each. The dried filters were placed in 4 ml of scintillation liquid, and the radioactivity was measured in a liquid scintillation counter (LS60001C, Beckman Coulter, USA). Radioactivity counts corresponding to incorporated thymidine, uridine, and methionine were normalized to OD450 at each time point. For the analysis of RFP expression relative to overall protein synthesis, the targeting and nontargeting cells were grown as described above, except that 500 μl-aliquots for 35S-methionine pulse-labeling reactions were taken before (time 0) and after RFP induction at 2, 5, and 10 min. The samples were analyzed by MOPS/SDS 10% polyacrylamide gel electrophoresis (PAGE) followed by Coomassie staining and quantification by Phosphorimager.

RNA isolation

Total RNA was isolated from E. coli cells growing in LB media and harvested at 60 min post-induction of RFP expression. Cell lysis was done using Max Bacterial Enhancement Reagent (Invitrogen) for 4 min and then with TRIzol reagent (Invitrogen) for 5 min. RNA was extracted by chloroform and precipitated with isopropanol. RNA pellets were washed with 70% ethanol and then dissolved in nuclease free water and then treated with Turbo DNA-free kit (Invitrogen) to remove DNA contamination.

Primer extension

For primer extension, 3.5 μg total RNA was reverse transcribed with the SuperScript IV First-Strand Synthesis System (Invitrogen) according to the manufacturer’s protocol. DNA oligonucleotides (Supplementary Table 2) were radiolabeled at the 5’ end by T4 Polynucleotide kinase (PNK) (New England Biolabs) treatment and γ-32P for 60 min at 37°C followed by purification on a Micro-Bio Spin P-6 Gel Column (Bio-Rad). 1 pmol of 5’-end-radiolabeled primer was used for each extension reaction. In parallel, sequencing reactions were performed on amplified PCR fragments of genomic or plasmid loci with the corresponding radiolabeled primers using the Thermo Sequenase Cycle Sequencing Kit according to manufacturer’s instructions. Reaction products were resolved by 10% denaturing PAGE and visualized using a Phosphorimager.

HEPN mutagenesis

Alanine mutations were created in each of four catalytic residues of LshCas13a HEPN domains (R597A, H602A, R1278A and H1283A) in plasmid containing LshCas13a locus and a CRISPR array carrying a spacer targeting RFP mRNA (pC003RFP1) using QuikChange Site-Directed Mutagenesis kit (Agilent) and the mutagenic primers containing the desired mutations (Supplementary Table 2) according to the manufacturer’s protocol. Cell growth experiments were done as described above. Primer extension analysis to determine cleavage products in _bla mRNA, rpmH mRNA and tRNAs was also done as described above. In all HEPN mutant experiments, wild-type targeting and nontargeting variants were used as controls.

Northern blot hybridization

The procedure was mostly performed as described before29. 10 μg of total RNA from E. coli cells transformed with CRISPR Cas13 plasmid expressing HEPN mutant LshCas13a proteins with mutations R597A, H602A, R1278A, H1283A, were resolved by 10% denaturing PAGE in Mini Protean 3 Cell (Bio-Rad). Separated RNA was then transferred to a nylon membrane (Hybond-XL, GE Healthcare) in pre-chilled 0.5xTBE buffer for 90 min at 50 V using Mini Trans-Blot electrophoretic transfer cell (Bio-Rad). Transferred RNA was UV-crosslinked to membrane and hybridized with γ- 32P 5’-end labeled RFP-spacer specific probe (RFP_crRNA_probe, Supplementary Table 2) in ExpressHyb solution (Clontech Laboratories, Inc) according to manufacturer’s protocol. Hybridization was performed in Isotemp rotisserie oven (Fisher Scientific) for 2 hours at 37°C. Hybridized RNA bands were visualized by phosphorimaging.

LshCas13a protein purification

Lsh cas13a gene was cloned into NdeI and BamHI sites in pET28 plasmid for protein purification. E. coli BL21(DE3) cells were transformed with pET28_LshCas13a for protein expression and isolation. LshCas13a protein was purified from the cells grown in 400 ml LB supplemented with 50 μg/ml kanamycin and 0.5 mM IPTG, as described previously30. Freshly transformed cells were grown at 37°C to OD600 0.6–0.9, then induced with 0.5 mM IPTG and grown for additional 6–8 hours at room temperature before cell harvesting. The cell pellets resuspended in buffer A (20 mM Tris-HCl pH, 8.0, 500 mM NaCl, 4 mM imidazole pH 8.0, 5% (v/v) glycerol, 0.2 μg/ml phenylmethylsulfonyl fluoride (PMSF)) containing protease inhibitor cocktail Roche cOmplete, EDTA-free (Sigma) were lysed by sonication. Lysates were cleared by centrifugation at 15,000 g for 60 min and filtered through 0.22 micron filter (Millipore) and applied to a 1-ml chelating Hi-Trap Sepharose column (GE Healthcare) equilibrated with buffer A. Proteins were purified first by washing the column using Buffer A containing 25 mM imidazole and then eluting with buffer A containing 200 mM imidazole. Pooled protein fractions were diluted 10 times with TGED buffer (20 mM Tris-HCl, pH 8.0, 5% (v/v) glycerol, 1 mM EDTA, 2 mM β-mercaptoethanol) and applied to a 1-ml Hi-Trap Heparin column (GE Healthcare) equilibrated with TGED. The column was washed with TGED containing 500 mM NaCl, and proteins were eluted with TGED containing 1 M NaCl. Pooled protein fractions were concentrated using Microsep centrifugal devices 30K (Pall Corp), dialyzed against buffer B (20 mM Tris-HCl, pH 8.0, 200 mM NaCl, 50% (v/v) glycerol, 0.5 mM EDTA, 2 mM β-mercaptoethanol) and then stored at −80°C. HEPN H602A encoding mutation was introduced into pET28_LshCas13a plasmid using QuikChange Site-Directed Mutagenesis kit as described above and mutant LshCas13a was purified similar to the wild-type LshCas13a.

Generation of RNA for in vitro cleavage assay

crRNA and target RNA were transcribed in vitro from PCR-generated dsDNA template using T7 RNA polymerase (New England Biolabs) according to manufacturer recommendations and was purified by electrophoresis in 10% polyacrylamide 6M urea gels. Oligonucleotide sequences are listed in Supplementary Table 2. Bulk E. coli tRNA was ordered from Sigma. Unmodified tRNAlys was ordered synthetically (Integrated DNA Technologies).

In vitro cleavage assay

In vitro RNA cleavage was performed with LshCas13a at 37 °C in cleavage buffer (20 mM Tris-HCl pH 8.0, 100 mM NaCl, 5 mM MgCl2, 1 mM DTT). LshCas13a-crRNA complex was formed by combining, in 10 μl, Cas13a and crRNA (200 nM each) and incubating at 37 °C for 20 min. Next, 100 nM of target RNA was added. Nontargeting control reactions were performed in the absence of target RNA. Immediately after adding target RNA, the reactions were supplemented with collateral RNA. 0.1 μg of E. coli bulk tRNA, 2 μg of total RNA or 0.05 μg of unmodified synthetic tRNAlys per 10 μl reaction were used as collateral cleavage substrates. After 60 min of incubation at 37 °C, RNA from reaction mixture was extracted with chloroform and precipitated in 75% ethanol in the presence of 0.3 M NaOAc and 0.1 mg/ml glycogen. tRNA cleavage products were analyzed by primer extension as described above. For RNA sequencing analysis of cleavage products generated by LshCas13a in vitro, total RNA from E. coli cells depleted from rRNA (MICROBExpress Bacterial mRNA Enrichment Kit (Invitrogen)) was used as cleavage substrate, the cleavage reactions were performed in 40 μl in triplicates.

RNA sequencing

The general procedure to prepare RNA for sequencing was similar to the protocol described previously31 with some modifications. Total RNA samples were treated with MICROBExpress Bacterial mRNA Enrichment Kit (Invitrogen) for rRNA depletion prior to library preparation. To construct libraries contained both primary transcripts carrying 5’-triphosphate (5’-P) and processed transcripts carrying 5’-monophosphate (5’-P) or 5’-hydrocyl (5’-OH), RNA samples were treated with RNA 5’ Pyrophosphohydrolase (RppH) (New England Biolabs) for 30 min at 37°C. In parallel, RNA samples without RppH treatment were used for preparation of libraries enriched in processed transcripts. Fragmentation was carried out by sonication using the Covaris protocol to obtain fragments of 200 nt. T4 PNK (New England Biolabs) treatment was performed to convert 5’-OH to 5’-P and 3’-P to 3’-OH and prepare RNA for adapter ligation during library preparation. Samples were purified using the Zymo Research Oligo Clean and Concentrator kit. Library preparation was done using the NEBNext Multiplex Small RNA Library Prep Set for Illumina according to the manufacturer’s protocol. BluePippin size selection was done using 2% agarose gel cassette (Sage Science) to select for 100 - 600 bp products. QC at each step was carried out by both Qubit and fragment analyzer. RNA-seq was performed using Illumina NextSeq High-Output kit 2 × 35 bp paired-end reads at Waksman Genomics Core Facility, Rutgers University. The similar procedure was performed for RNA library preparation and sequencing of RNA cleavage products produced by target-activated LshCas13a in vitro. RNA-seq _in vit_ro cleavage products was performed at Skoltech Genomics Core Facility.

RNA sequencing data analysis

All custom scripts used in the analysis are deposed at GitHub:

Raw RNA sequencing reads were filtered by quality with simultaneous adapters removal using trimmomatic v. 0.3632. The exact parameters of trimmomatic run are available in the file. Adapters content and quality of reads before and after the processing was assessed using FastQC v. 0.11.9. Processed reads were mapped onto reference sequences (RefSeq: NC000913.3 supplemented with pC002 and pC008 plasmids for nontargeting samples and NC_000913.3 supplemented with pC003_RFP_spacer and pC008 plasmids for targeting samples), using bowtie2 v. producing corresponding SAM files. For the transcription start sites (TSS) detection, processed reads for nontargeting samples with or without RppH treatment were mapped to the NC_000913.2 sequence. To analyze RNA-seq data for _in vitro cleavage experiments, RNA reads were mapped to the NC000913.3 sequence. The exact parameters of bowtie2 run are available in the file. Next, for each nucleotide position of each strand of reference sequences, the number of 5’ ends of aligned fragments were counted, producing corresponding tables (see file for details). The obtained tables were joined using script. The differences between the numbers of mapped 5’ ends in targeting and nontargeting samples were analyzed using edgeR package v. 3.26.334. The features (here, strand specific nucleotide positions) with low counts were excluded from the analysis. The trimmed mean of M-values (TMM) normalization method implemented in edgeR was applied. Next, the edgeR likelihood ratio test was performed. The obtained p-values were corrected using Benjamini-Hochberg method, and the result tables containing analyzed features with assigned log2FC and adjusted p-values were written to separate files (see TCS_calling.R file for details). The features with log2FC > 0 were considered as putative RNA cleavage sites. To build weblogo plots, the identified RNA cleavage sites were sorted by adjusted p-values in ascending order, and top-1000 were selected for the analysis. Ten nucleotides surrounding the selected sites were obtained from the reference sequences using Biopython toolkit35, and the Logomaker36 module was used to create weblogo plots (see TCS_LRT_weblogo.ipynb file for details). Secondary structures of transcripts were predicted using the RNAfold tool from the ViennaRNA package37 and visualized using the forgi module38 (see draw_hairpin_structures.ipynb file for details). To visualize the cleavage of different tRNAs, the sequences of tRNA genes that were included in the analysis, were extracted from annotated NC000913.3 assembly and aligned using MAFFT v. 7.453 with --maxiterate 1000 --localpair parameters. Each position of the alignment was assigned with the corresponding -log10(adjusted p-value) depicting the statistical significance of the enrichment of 5’ ends counts in targeting samples over nontargeting samples in d10 strain. The positions corresponding to gaps or the positions excluded from the analysis were assigned with zero values. The resulting table was visualized as the heatmap where the intensity of color depicts the -log10(adjusted p-values). To validate the approach used for identification of RNA cleavage sites, we apply this method to define TSS. The differences between the numbers of mapped 5’ ends in the samples with and without RppH treatment were analyzed in the same way as it was done for the detection of RNA cleavage sites. Predicted TSSs were compared with the _E. coli MG1655 TSSs list from RegulonDB39.

Supplementary Information is available and contains a Supplementary Discussion, Supplementary Tables 1 and 2, Supplementary Figures 1 and 2, gel source data, and Supplementary analysis.

Article TitletRNA anticodon cleavage by target-activated CRISPR-Cas13a effector


Type VI CRISPR-Cas systems are the only CRISPR variety that cleaves exclusively RNA1,2. In addition to the CRISPR RNA (crRNA)-guided, sequence-specific binding and cleavage of target RNAs, such as phage transcripts, the type VI effector, Cas13, causes collateral RNA cleavage, which induces bacterial cell dormancy, thus protecting the host population from phage spread3,4. We show here that the principal form of collateral RNA degradation elicited by Cas13a protein from Leptotrichia shahii upon target RNA recognition is the cleavage of anticodons of multiple tRNA species, primarily those with anticodons containing uridines. This tRNA cleavage is necessary and sufficient for bacterial dormancy induction by Cas13a. In addition, Cas13a activates the RNases of bacterial toxin-antitoxin modules, thus indirectly causing mRNA and rRNA cleavage, which could provide a back-up defense mechanism. The identified mode of action of Cas13a resembles that of bacterial anticodon nucleases involved in antiphage defense5, which is compatible with the hypothesis that type VI effectors evolved from an abortive infection module6,7 encompassing an anticodon nuclease.

Login or Signup to leave a comment
Find your community. Ask questions. Science is better when we troubleshoot together.
Find your community. Ask questions. Science is better when we troubleshoot together.

Have a question?

Contact or check out our support page.