Addressing the objectives defined for the NascenTomiX project required us to develop new methodologies. We first established iTP-seq, a scalable and versatile tool to identify peptide-encoding transcripts that induce translational arrest in vitro (Seip et al., 2018). While it is similar to Ribo-seq, a technique that is widely used to monitor patterns of gene expression in living cells, iTP-seq can be used on transcript libraries of any size, composition or complexity, making it ideal to analyze translational landscapes at a fraction of the cost.
We have used iTP-seq to study how bacteria harboring an ermD resistance gene under the control of an ErmDL arrest peptide become resistant to macrolides upon exposure to these antibiotics (Beckert et al., 2021). Coupled to a study on the drug-dependent arrest peptide MsrDL (Fostier et al., 2023), our work shed light on the mechanisms used by certain bacteria to control the expression of resistance genes. Moreover, we found that iTP-seq is ideally suited to the study of ribosome-targeting antibiotics, such as the aromatic polyketide Tetracenomycin X (Leroy et al., 2023). Identifying detailed mechanisms of action for widely-used, recently discovered or forgotten antibiotics should facilitate the development of improved therapeutics.
In order to identify naturally occurring arrest peptides in bacteria, we developed retapamulin-assisted inverse toeprinting (RET-iTP), a genome-wide profiling method for the identification of translation initiation sites in vitro that is independent of transcript abundance (to be published). During the course of the project, we discovered and characterized SpeFL, an arrest peptide that functions as an L-ornithine sensor to activate polyamine biosynthesis in γ-proteobacteria (Herrero del Valle et al., 2020). Using a combined structural and biochemical approach, we determined that activation of the putrescine biosynthesis gene speF relies upon ribosome stalling resulting from the capture of L-ornithine by a ribosome translating speFL. Moreover, we showed structurally how the ribosome and SpeFL form a highly selective binding pocket for L-ornithine. Similarly, we determined the mechanism by which the arrest peptide TnaC senses the amino acid L-tryptophan to trigger indole production in γ-proteobacteria (van der Stel et al., 2021). Together, these studies revealed the basic principles underlying the detection of small molecules with low intrinsic affinity for the ribosome by arrest peptides.
Although iTP-seq is a key tool in our search for metabolite-dependent arrest peptides, evolving peptides that target ribosomes in trans required us to develop a different kind of methodology. We therefore devised a droplet-based microfluidics approach that compartmentalizes an in vitro translation reaction, such that a single peptide variant is produced per droplet (to be published). The effect of each peptide on the production of a fluorescent reporter inside live bacteria is monitored and fluorescence-activated droplet sorting is used to retain droplets that contain inhibitory peptides. We have established that this approach can be used to select peptides that inhibit translation and plan to incorporate it into a novel pipeline to develop ribosome-targeting antimicrobial peptides.
