An archaeal CRISPR type III-B system exhibiting distinctive RNA targeting features and mediating dual RNA and DNA interference

Abstract

CRISPR-Cas systems provide a small RNA-based mechanism to defend against invasive genetic elements in archaea and bacteria. To investigate the in vivo mechanism of RNA interference by two type III-B systems (Cmr-α and Cmr-β) in Sulfolobus islandicus, a genetic assay was developed using plasmids carrying an artificial mini-CRISPR (AC) locus with a single spacer. After pAC plasmids were introduced into different strains, Northern analyses confirmed that mature crRNAs were produced from the plasmid-borne CRISPR loci, which then guided gene silencing to target gene expression. Spacer mutagenesis identified a trinucleotide sequence in the 3'-region of crRNA that was crucial for RNA interference. Studying mutants lacking Cmr-α or Cmr-β system showed that each Cmr complex exhibited RNA interference. Strikingly, these analyses further revealed that the two Cmr systems displayed distinctive interference features. Whereas Cmr-β complexes targeted transcripts and could be recycled in RNA cleavage, Cmr-α complexes probably targeted nascent RNA transcripts and remained associated with the substrate. Moreover, Cmr-β exhibited much stronger RNA cleavage activity than Cmr-α. Since we previously showed that S. islandicus Cmr-α mediated transcription-dependent DNA interference, the Cmr-α constitutes the first CRISPR system exhibiting dual targeting of RNA and DNA.

Figure 1.

Plasmid-borne artificial mini-CRISPR locus-mediated gene silencing in S. islandicus. (A) Schematic of three sequence regions selected for studying gene silencing of lacS expression. Sequences of the antisense strand from 651 to 690 and 1169–1208 constitute protospacer 1 (ProtoS1) and ProtoS3, respectively, whereas ProtoS2 is the sense sequence from 703 to 742 nt, relative to the start codon (ATG, +1) of the lacS gene. The pentanucleotides (5′-GAAAG-3′ or 5′-GAGAC-3′) are underlined. Ter, terminator of the lacS gene. (B) Expression of mini-CRISPR loci from pAC-SS1 and pAC-SS2 and crRNA maturation. Northern analysis of total RNAs prepared from cells of S. islandicus E233 transformants harbouring pSeSD1, pAsRNA, pAC-SS1 and pAC-SS2. Probes used in these analyses are indicated individually (S1, S2, or A2S32). The two former detect crRNAs generated from the mini-CRISPR loci on plasmid whereas the latter identifies mature crRNAs generated from spacer 32 in the genomic CRISPR locus 2 (A2S32) of S. islandicus E233S1. 5s rRNA was used as a loading control. T – RNA transcripts; PI – processing intermediates; crRNAs – mature crRNAs. (C) Specific β-glycosidase activity in these transformants. Three independent transformants were analysed for each construct with bars indicating standard deviations.

Figure 2.

Identification of a trinucleotide sequence motif in crRNA important for RNA interference. Complete sequence of S1 is shown on the top of the left panel and numbered. S1 spacer mutants are annotated according to the positions of their mutations in the spacer and the mutated bases are shown in the left panel. The right panel illustrates RNA targeting activity of the wild-type S1 spacer (100%) and mutated derivatives relative to the wild-type spacer. Spacer mutants showing a greatly reduced RNA targeting activity are highlighted in colours. Three independent transformants were analysed for each construct with bars indicating standard deviations.

Figure 3.

Each type III-B Cmr system mediated gene silencing in S. islandicus. (A) Schematic of pAC-SS1-LacS, an RNA interference test plasmid producing both S1 crRNA and its RNA substrate. After introducing pAC-SS1-LacS into S. islandicus E233S1 or mutants of cmr gene cassettes by transformation, mature crRNAs were to be produced from the SS1 mini-CRISPR locus and then guided the endogenous Cas proteins to silence lacS mRNAs, which were also produced from the lacS gene on the plasmid (3–5 copies per cell). (B) Specific β-glycosidase activity in S. islandicus Cmr mutants. Cmr-αβ – the wild-type strain carrying both cmr gene cassettes, Cmr-α – mutant retaining Cmr-α with cmr-β gene cassette deleted, Cmr-β – mutant retaining Cmr-β with cmr-α genes deleted, and Cmr-null – mutant with both cmr gene cassettes deleted. S. islandicus E233S1 strain carrying pLacS-ck was used as a control. Three independent transformants were analysed for each construct. Bars: standard deviations.

Figure 4.

Target RNA cleavage analysis by real-time qPCR and 3′-RACE. (A) Strategy for detecting RNA cleavage at the S1 protospacer in mRNA. Qtar primer set amplified a region containing the predicted RNA cleavage sites while Qref amplified a region positioned upstream the S1 protospacer in mRNA. (B) Quantification of mRNA cleavage in RNA targeting strains of S. islandicus E233 carrying the chromosomal target gene lacS. Total RNAs were prepared from S. islandicus E233 cells harbouring pAC-SS1 or pSeSD1 and used for real-time qPCR. The amount of uncut mRNA is expressed as a ratio of lacS mRNA levels in transformants of pSeSD1 versus pAC-SS1. Bars represent the standard deviations of triplicates. (C) Quantification of mRNA cleavage in RNA targeting strains derived from S. islandicus E233S1 carrying the plasmid-borne lacS. Strains used are the same as in Figure 3B. Uncut mRNA is calculated as the percentage of the mRNA level in each strain carrying pAC-SS1 and that in S. islandicus E233S1 containing pLacS-ck. Bars represent the standard deviations of triplicates. (D) Schematic of cleavage sites within the S1 protospacer identified by 3′-RACE. Both protospacer S1 and the crRNA are shown with positions are numbered on crRNA. On RNA target: protospacer is shown as capital letters whereas the flanking nucleotides are in lowercase letters. The repeat handle of crRNA appears in green. Cleavage sites are indicated with arrows. Red arrows indicate cleavage sites by Cmr-α while Black arrows show cleavage sites by Cmr-β. Numbers below arrows indicate PCR clones carrying each cleavage site.