Dual modes of CRISPR-associated transposon homing

Abstract

Tn7-like transposons have co-opted CRISPR systems, including class 1 type I-F, I-B, and class 2 type V-K. Intriguingly, although these CRISPR-associated transposases (CASTs) undergo robust CRISPR RNA (crRNA)-guided transposition, they are almost never found in sites targeted by the crRNAs encoded by the cognate CRISPR array. To understand this paradox, we investigated CAST V-K and I-B systems and found two distinct modes of transposition: (1) crRNA-guided transposition and (2) CRISPR array-independent homing. We show distinct CAST systems utilize different molecular mechanisms to target their homing site. Type V-K CAST systems use a short, delocalized crRNA for RNA-guided homing, whereas type I-B CAST systems, which contain two distinct target selector proteins, use TniQ for RNA-guided DNA transposition and TnsD for homing to an attachment site. These observations illuminate a key step in the life cycle of CAST systems and highlight the diversity of molecular mechanisms mediating transposon homing.

Keywords: CRISPR, CRISPR-associated transposases, CAST, RNA-guided DNA transposition, homing transposition, Type V-K Cas12k effector, Type I-B Cascade effector, Tn7, TnsD/TniQ, transposon target selectors.

Figure 1|. Homing of type V-K ShCAST is guided by a short, delocalized crRNA.

(A) Representative CAST systems from 57 analyzed loci with their respective homing sites in yellow. (B) Diagram of the delocalized crRNA and corresponding spacer match in downstream tRNA-Leu from S. hofmannii CAST. (C) ShCAST homing into a tRNA target plasmid in E. coli. Schematic of the natural delocalized crRNA and insertion activity of crRNA variants; promoters and terminators are indicated with an arrow and hairpin, respectively. Data are represented as mean ± SD. (D) Deep sequencing of the ShCAST homing insertion product. (E) Schematic representation of the ShCAST tracrRNA interacting with a direct repeat from the canonical CRISPR array.

Figure 2|. Diversity of delocalized crRNA across type V-K CAST systems.

(A) Bioinformatic analysis of sequence and distance from the typical CRISPR array of the delocalized crRNA in five CAST systems (See Supplementary Table 1 for full list). (B) Two type V-K CAST systems (Halothece sp. PCC 7418 and Cyanobacterium sp. HL-69) that contain a short crRNA targeting the adjacent tRNA, but no detectable CRISPR array. (C) Position of the homing target PAM and corresponding left end (LE) for tRNA-homing CAST systems. (D) RNA sequencing reads from S. hofmannii aligned to the delocalized crRNA locus. (E) RNA sequencing reads from A. cylindrica aligned to the delocalized crRNA locus.

Figure 3|. CAST I-B loci architecture and comparison of the TniQ/TnsD target selectors.

(A) Schematic of the Anabaena variabilis ATCC 29413 CAST I-B subtype 1 (I-B1) locus and ‘Peltigera membranacea cyanobiont’ 210A CAST I-B subtype 2 (I-B2) locus containing both Cascade proteins, Tn7-like proteins including TnsD and TniQ target selectors (Figure S3a, S3b. S3c). (B) A dendrogram showing the similarity of TniQ/TnsD proteins of Tn7 and CAST I-B1 and I-B2, CAST I-F, and CAST V-K systems (Figure S3b, S3d). Dark purple indicates a similar N-terminal region (TniQ/TnsD core). Light purple indicates similarity between the C-terminal regions of Tn7-TnsD, CAST I-B1 TnsD, and remote similarity with CAST I-B2 TnsD (see Figure S4, S12). Dashed lines indicate boundaries of TniQ/TnsD core. Light and dark yellow colors indicate the protein domains annotated from the CAST I-F TniQ structure (see Figure S3d). (C) Phylogenetic tree built with FastTree (WAG model) (Price et al., 2009) from the core region (Figure S3d) of TniQ/TnsD proteins from CAST I-B (TniQ in red; TnsD in purple), Tn7-TnsD (in green), CAST V-K TniQ (in yellow), and CAST I-F TniQ (in cyan). The tree indicates proximity between CAST I-B1 TnsD and Tn7-TnsD and the distinction between CAST I-B1 and CAST I-B2 (Figure S3b, S3d).

Figure 4|. Characterization of Type I-B subtype1 AvCAST.

(A) Schematic of experiment to identify insertion direction, PAM, and insertion position by AvCAST in E.coli. (B) RNA sequence reads mapping to the minimal CRISPR array to reveal the mature crRNA sequence of AvCAST. (C) Insertion directionality assayed by diagnostic PCR of pInsert on 6N PAM library plasmid with different primer pairs in ΔTniQ or ΔTnsD conditions. (D) Quantification of insertion frequency in both directions by ddPCR. Data are represented as mean ± SD. (E) Top. PAMs for AvCAST RNA-guided insertions. Bottom. AvCAST RNA-guided insertion positions identified by deep sequencing. (F) Long-read Nanopore sequencing to characterize the structure of pInsert. (G) Sanger sequencing chromatograms of a representative pInsert. PAM, AvPSP1, TSDs, and transposon ends are annotated.

Figure 5|. Homing of AvCAST is mediated through conserved attTn7 recognition by TnsD.

(A) RNA-guided insertion frequency of AvCAST into the pTarget with AvPSP1 and glmS gene at each ΔTniQ/ΔTnsD condition. Data are represented as mean ± SD. (B) Tn7-like machinery-mediated insertion frequency into the pTarget with AvPSP1 and glmS gene at each ΔTniQ or ΔTnsD condition. Data are represented as mean ± SD. (C) AvCAST-mediated single prominent insertion at glmS Tn7 attachment site on plasmid identified by deep sequencing. Purple bar indicates the position by Tn7-like machinery (tnsA, tnsB, tnsC, and tnsD). Blue bar indicates the position by all AvCAST protein components (Tn7-like machinery + TniQ + Cascade). Light blue bar indicates the insertion position on the plasmid harboring an additional 65-bp downstream sequence of A. variabilis glmS. (D) Tn7-like machinery-mediated insertion frequency of AvCAST into a plasmid bearing the A.variabilis glmS gene, a plasmid bearing mutations in the conserved attTn7 site, and a plasmid bearing the E.coli glmS gene. The 30 bp of the C-terminus of glmS, a part of attTn7 identified by a previous study (Mitra et al.), is shown. Base numbering format follows that of the study (End of glmS = +23), and reported essential positions for TnsD recognition are shown in red. Data are represented as mean ± SD. (E) Schematic of in vitro transposition reactions with purified Tn7-like machinery components of AvCAST. (F) Tn7-like machinery and donor requirements for in vitro transposition on E. coli glmS Tn7 attachment site. pInsert was detected by PCR for LE and RE junctions. (G) Tns protein requirements for in vitro transposition to the E.coli glmS Tn7 attachment site. All reactions contained pDonor and pTarget.

Figure 6|. Type I-B subtype 2 PmcCAST system homes to tRNA-val.

(A) Top. PAMs for PmcCAST RNA-guided insertions. Bottom. AvCAST RNA-guided insertion positions identified by deep sequencing. (B) Long-read Nanopore sequencing to characterize the structure of pInsert. (C) Schematic of experiment to compare RNA-guided transposition and homing to tRNA-val by PmcCAST in E.coli. (D) RNA-guided insertion frequency of PmcCAST into the pTarget with PmcPSP1 and tRNA-val gene at each ΔTniQ or ΔTnsD condition. Data are represented as mean ± SD. (E) Tn7-like machinery-mediated insertion frequency of PmcCAST into pTarget at each ΔTniQ or ΔTnsD condition. Data are represented as mean ± SD. (F) PmcCAST-mediated prominent insertion at tRNA-val gene on target plasmid identified by deep sequencing. Purple bar indicates the position by all Tns proteins (TnsAB, TnsC, TniQ, and TnsD). Blue bar indicates the position at ΔTniQ condition. Light blue bar indicates the position at ΔTnsD condition. White bar indicates the position at ΔTniQΔTnsD conditions. (G) Tn7-like machinery and donor requirements for in vitro transposition on tRNA-val gene. pInsert was detected by PCR for LE and RE junctions. (H) Tns protein requirements for in vitro transposition on tRNA-val gene. All reactions contained pDonor and pTarget.

Figure 7|. Models of transposition mechanisms of CAST I-B and CAST V-K systems.

(A) Prototypical loci organization of CAST I-B and CAST V-K systems. The loci are delimited by transposon ends (Tn ends, indicated as light vertical rectangle). Transposon core components (tnsA, tnsB, and tnsC in CAST I-B; tnsB and tnsC in CAST V-K) are colored in light blue while tniQ are indicated in purple with a lighter purple for tnsD. Cas components (Cascade and cas12k) are shown in grey. Conserved transcription factors (TF) are shown in pink. Locations of cargo genes are indicated with dark grey circles. CRISPR arrays are shown as grey triangles for the repeats and red diamonds for the spacers. The delocalized crRNA is indicated by a lighter triangle (partial repeat) and a truncated red diamond (short spacer). Homing target sites are colored in yellow. Horizontal light grey triangle indicates the tracrRNA in the CAST V-K locus. Conserved transcription factors (TF) are shown in light pink. (B) Models for CAST transposition to mobile element (donor cell in blue) and CAST homing transposition to bacterial chromosome (recipient cell in green). CAST transposition to mobile genetic elements is mediated by the Cas effector machinery (IB Cascade and crRNA or Cas12k, tracrRNA, and cRNA; Cas proteins are shown in grey) that recognizes the target site (red) with RNA-guided targeting to insert the transposon using Tn components. TniQ may function as an adaptor between Cas machinery and the transpososome (Tn core machinery bound to the DNA transposon). The mobile genetic element (MGE) where the transposon has inserted can be horizontally transferred (via HGT) to another host where homing transposition can occur. CAST I-B homing transposition is mediated by TnsD (without Cas components), which targets the homing site (light orange) located in the bacterial chromosome (in green). CAST V-K uses a dedicated crRNA to target the homing site.