Structural and functional characterization of a bacterial TenpIN type III toxin-antitoxin system and its potential as an antibacterial strategy

Abstract

Bacterial type III toxin-antitoxin (TA) systems were first identified as abortive infection systems wherein a phage infection causes altruistic suicide of the bacterium to prevent bacteriophage spread among its clonal population. Type III TA systems consist of a cognate antitoxin RNA to counteract its toxin protein under homeostatic conditions. These TA systems are classified into tenpIN, cptIN, and toxIN families, based on the sequence homology of the toxin and antitoxin. Here, we identified 722 unique putative tenpIN (for Type III ENdogenous to Photorhabdus Inhibitor/toxIN) homologues, the least studied family, in several pathogens and showed that it is wide-spread among prokaryotes. We further investigated these systems through three case studies involving ESKAPE pathogens, E. coli, and viruses. Our analysis revealed operon organization patterns and conserved residues within TenpN proteins, which we mapped onto predicted AlphaFold structures. Additionally, our bioinformatics study showed that the antitoxin RNAs possess conserved regions across bacterial species and adopt complex secondary structures. Furthermore, we have functionally characterized the first TenpIN system from E. coli, carried out large-scale purification of the RNP complex, and solved its structure by single-particle cryo-electron microscopy (cryoEM) and de novo modelling at a 3.9 Å resolution. The cryoEM map showed that it consists of a hetero-tetramer consisting of two proteins and two RNA moieties forming a closed assembly. The RNA forms a pseudoknot structure with the single stranded region on either end interacting with the active site of the toxin protein which is an endoribonuclease. The core pseudoknot seems to be important for the fold and stability of the RNP complex. Additionally, we assessed bacterial cell morphology via microscopy and toxin’s downstream targets through transcriptomics without the expression of the cognate antitoxin RNA. Interestingly, we observed that these cells were severely damaged and were unable to grow further in the absence of the antitoxin RNA. Under this stress condition, the bacteria are more susceptible to antibiotics as compared to empty vector controls. These observations provide a platform to design agents that activate these systems in bacterial cells as a potential antibacterial strategy.