Specificity and mechanism of tRNA cleavage by the AriB Toprim nuclease of the PARIS bacterial immune system

Abstract

Transfer RNA (tRNA) molecules have been recently recognized as widespread targets of bacterial immune systems. Translation inhibition through tRNA cleavage or modification inhibits phage propagation, thereby protecting the bacterial population. To counteract this, some viruses encode their own tRNA molecules, allowing infection to take place. The AriB effector of the PARIS defence system is a Toprim nuclease previously shown to target the Escherichia coli tRNA^Lys(UUU), but not a tRNA^Lys(UUU) variant encoded by bacteriophage T5. We demonstrate here that the T5 tRNA^Lys(UUU) is required but not sufficient to bypass PARIS immunity. Combining tRNA sequencing, genetics, phage infection and in vitro biochemical data, we reveal that the E. coli tRNA^Thr(UGU) is another prime target of AriB, and tRNA^Asn(GUU) represents a secondary, yet biologically relevant, target of the PARIS effector. Activated AriB protein cleaves these targets in vitro, and the cleavage reaction is not dependent on the presence of specific tRNA modifications. We show that the overexpression of phage T5 tRNA^Lys(UUU), tRNA^Thr(UGU) and tRNA^Asn(GUU) variants is sufficient to inhibit PARIS anti-viral defence. Finally, we propose a model for tRNA recognition by the AriB dimer and provide molecular details of its nuclease activity and specificity.This article is part of the discussion meeting issue 'The ecology and evolution of bacterial immune systems'.

Keywords: PARIS; bacterial immunity; bacteriophage; nuclease; tRNA.

Figure 1.

Landscape of AriB-mediated tRNA cleavages. (A) Plaque assays were conducted using cells without (green fluorescent protein (GFP) control) or with PARIS in the presence of empty vector (pBAD) or pFD287 (paraBAD T5 tRNA^Lys(UUU)) against T5_WT, T5_Mos and T5₁₂₃. Overexpression of T5 tRNA^Lys(UUU) in PARIS cells from pFD287 plasmid, with the tRNA^Lys(UUU) of phage T5 under the control of the araBAD promoter inducible by L-Ara (0.2%), rescues infection by phage T5₁₂₃ but not by T5_Mos. (B–D) Scatter plots of normalized tRNA abundances. Axes represent the normalized abundance of tRNAs in the respective samples. Blue dots indicate the majority of tRNAs, while red dots highlight specific tRNAs of interest: Lys(UUU), Thr(UGU), Thr(CGU) and Asn(GUU). The dashed grey line shows the linear regression fit in log–log space. (B) tRNAs with mapped reads ≥65 nt from GFP control and PARIS (in vivo cultures), sequenced using a custom protocol (see §2). (C) Small tRNAs (<45 nt) with 5′ ends mapped to the anticodon stem–loop, from GFP control and PARIS (in vitro cultures), sequenced using a commercial kit. (D) Small tRNAs (<45 nt) with 5′ ends mapped to the anticodon stem–loop, from untreated GFP control and in vitro AriB cleavage samples, sequenced using a commercial kit. (E,F) Frequency of ribosome pausing on specific codons, estimated by toeprinting analysis with substrates RST3_all_tRNAs (E) and RST3_all_tRNAs (K5A) (F) mRNA templates translated in the presence of AriB wild-type (WT) or E26A Toprim mutant. Data are presented as relative fluorescence units (rfu). Sequences of mRNA templates are provided at the top. ‘Ths’ represents a control reaction with thiostrepton, which stalls ribosomes at the start codon. TAKM7 is a standard control buffer for in vitro translation. Signals from the start (AUG), lysine (AAA) and threonine (ACA) codons were calculated as the mean area under the curve from the capillary electropherograms carried out in triplicates; representative electropherograms are presented in electronic supplementary material, figure S1D,E. DMSO, dimethyl sulfoxide.

Figure 2.

Expression of viral tRNAs reduces AriB toxicity and restores phage infectivity. (A) Phage T5 tRNA^Lys(UUU)/tRNA^Thr(UGU)/tRNA^Asn(GUU) expression can rescue PARIS toxicity triggered by the expression of the phage T7 Ocr from 2,4-diacetylphloroglucinol (DAPG)-inducible PhlF promoter. Top: without the PARIS system. Bottom: with the PARIS system. (B) Heatmap of PARIS defence against a panel of phages. Colour intensity indicates plaque-forming units (PFUs) on a log₁₀ scale. The panel below shows whether phages encode their own variant of the tested tRNAs. Representative plates used to build the heatmap as well as error bars are shown in electronic supplementary material, figure S2B,C.

Figure 3.

AriB does not require tRNA base modifications for cleavage. (A) AriB toxicity was triggered with 2,4-diacetylphloroglucinol (DAPG, 50 µM) to induce the expression of T7 Ocr under the DAPG-inducible PPhlF promoter in Escherichia coli (Ec) strains harbouring different deletions of genes implicated in base modifications with either a green fluorescent protein (GFP) control vector (A) or the PARIS system (B). In vitro AriB cleavage assay on tRNAs extracted from the four deletion strains or the wild-type (WT) strain. (C) Activity of AriB on specific in vitro transcripts corresponding to E. coli tRNAs. (D) Activity of AriB on E. coli (Eco) and T5 tRNA^Lys(UUU) and mutants; point mutations are shown in electronic supplementary material, figure S3D.

Figure 4.

Predicted structure and mechanism of AriB. (A) Overall structure of AriB dimer bound to Escherichia coli tRNA^Lys(UUU) generated with AlphaFold 3 and showing the monomers AriB and AriB’ in light blue and light green. The cleavage site is marked by a magenta sphere. (B) Left: view of the electrostatic surface of the AriB dimer bound to the E. coli tRNA^Lys(UUU), showing the two lysines K60and K64 stabilizing the anticodon loop on the AriB monomer with the active catalytic site. Right: mutations predicted to stabilize the anticodon loop (K60A, K64A) prevent PARIS-mediated defence. Efficiency of plaquing of phage T7 on E. coli MG1655 carrying the wild-type (wt) or mutated PARIS system. (C) Superposition of the M5 active site with the AriB active site in complex with the E. coli tRNA^Lys(UUU). (D) Active catalytic site of AriB represented with two Mg²⁺ ions modelled. (E) Incubation of AriB cleavage products with poly(A) polymerase run on a 12% denaturing gel: lane L, ssRNA ladder; lane 1, 100 nM transcript tRNA^Lys(UUU); lane 2, same as lane 1, + 25 nM AriB dimer + 1 mM Mg²⁺; lane 3, same as lane 2, with poly(A) polymerase treatment post cleavage.