Biology subjects

Mordret, E.

Publications and source records attributed to Mordret, E..

6 recordsLinked to original sources

bioRxiv2025

Protein and genomic language models chart a vast landscape of antiphage defenses

The bacterial pangenome encodes an immense array of antiphage systems, yet much of their diversity remains uncharted. In this study, we developed language models to predict novel antiphage proteins in two ways: first via fine-tuning ESM2, a protein language model capable of detecting distant homology to known defense proteins, second via a genomic language model with ALBERT architecture which predicts defensive function based on genomic context. We demonstrate that applying these approaches to Actinomycetota - a phylum largely unexplored for antiphage defenses, can accurately predict previously unknown functional defense mechanisms, leading to the discovery and experimental validation of six defense systems with novel antiphage proteins. Analysis of over 30,000 bacterial genomes predicted more than 45,000 uncharacterized protein families potentially involved in antiphage defense, underscoring the vast, untapped diversity of these systems.

microbiology↗

bioRxiv2024

A family of lanthipeptides with anti-phage function

Bacteria synthesize natural products to adapt to their environment, where phage-bacteria interactions play a crucial role in bacterial ecology. Although a few natural products have been shown to protect bacteria from phage infection, the prevalence and diversity of chemical anti-phage defense remain largely unexplored. Here, we uncover a novel family of over 2000 lanthipeptide biosynthetic gene clusters from Actinobacteria that participate in anti-phage defense, which we named lanthiphages. Lanthiphages colocalize with other anti-phage systems in defense islands. We demonstrate that native lanthiphage expression protects the model strain Streptomyces coelicolor against diverse phages. Heterologous expression of four additional lanthiphage pathways shows that the anti-phage function is conserved across this family of biosynthetic gene clusters. Finally, we demonstrate that lanthiphage expression leads to the production of a novel compound and alters phage transcription. Our findings highlight that biosynthetic gene clusters with anti-phage functions can be successfully identified through genomic analysis. This work paves the way for the systematic mining of anti-phage natural products, which could constitute a novel reservoir of antiviral drugs.

microbiology↗

bioRxiv2024

A Comprehensive Resource for Exploring Antiphage Defense: DefenseFinder Webservice, Wiki and Databases.

In recent years, a vast number of novel antiphage defense mechanisms were uncovered. To facilitate the exploration of mechanistic, ecological, and evolutionary aspects related to antiphage defense systems, we released DefenseFinder in 2021 (Tesson et al., 2022). DefenseFinder is a bioinformatic program designed for the systematic identification of known antiphage defense mechanisms. The initial release of DefenseFinder v1.0.0 included 60 systems. Over the past three years, the number of antiphage systems incorporated into DefenseFinder has grown to 152. The increasing number of known systems makes it a challenge to enter the field and makes the interpretation of detections of antiphage systems difficult. Moreover, the rapid development of sequence-based predictions of structures offers novel possibilities of analysis and should be easily available. To overcome these challenges, we present a hub of resources on defense systems, including: 1) an updated version of DefenseFinder with a web-service search function, 2) a community-curated repository of knowledge on the systems, and 3) precomputed databases, which include annotations done on RefSeq genomes and structure predictions generated by AlphaFold. These pages can be freely accessed for users as a starting point on their journey to better understand a given system. We anticipate that these resources will foster the use of bioinformatics in the study of antiphage systems and will serve the community of researchers who study antiphage systems. This resource is available at: https://defensefinder.mdmlab.fr.

genomics↗

bioRxiv2022

Conservation of antiviral systems across domains of life reveals novel immune mechanisms in humans

Viral infection is a common threat to prokaryotic and eukaryotic life, which has resulted in the evolution of a myriad of antiviral systems. Some of these eukaryotic systems are thought to have evolved from prokaryotic antiphage proteins, with which they may display sequence and structural homology. Here, we show that homologs of recently discovered antiphage systems are widespread in eukaryotes. We demonstrate that such homologs can retain a function in immunity by unveiling that eukaryotic proteins of the anti-transposon piRNA pathway display domain homology with the antiphage system Mokosh. We further utilise this conservation to discover novel human antiviral genes related to the Eleos and Lamassu prokaryotic systems. We propose that comparative immunology across domains of life can be leveraged to discover immune genes in eukaryotes.

genomics↗

bioRxiv2018

Extensive programmed ribosomal frameshifting in human as revealed by a massively parallel reporter assay

Programmed ribosomal frameshifting is the controlled slippage of the translating ribosome to an alternative frame. This tightly regulated process is widely employed by human viruses such as HIV and SARS coronavirus and is critical for their life cycle and virulence. It is also utilized from yeast to human to implement a feedback control mechanism to regulate polyamine levels. Here, we developed a high-throughput, fluorescence-based approach to assess the frameshifting potential of a sequence. We designed and tested >12,000 sequences based on 15 viral and human frameshifting events, allowing us to elucidate the rules governing ribosomal frameshifting in a systematic way and to discover novel regulatory inputs based on amino acid properties and tRNA availability. We assessed the natural variation in HIV gag-pol frameshifting rates by testing >500 clinical isolates and identified subtype-specific differences as well as associations between viral load in patients and the optimality of gag-pol frameshifting rates. We further devised computational models that accurately predict frameshifting potential (up to auROC=0.93) and frameshifting rates (up to Pearson r=0.81) of novel variants, including subtle differences between HIV clinical isolates (r=0.60). Taken together, this systematic approach can contribute to the development of antiviral agents acting on programmed ribosomal frameshifting.

genomics↗

bioRxiv2018

Systematic detection of amino acid substitutions in proteome reveals a mechanistic basis of ribosome errors

Translation errors limit the accuracy of information transmission from DNA to proteins. Selective pressures shape the way cells produce their proteins: the translation machinery and the mRNA sequences it decodes co-evolved to ensure that translation proceeds fast and accurately in a wide range of environmental conditions. Our understanding of the causes of amino acid misincorporations and of their effect on the evolution of protein sequences is largely hindered by the lack of experimental methods to observe errors at the full proteome level. Here, we systematically detect and quantify errors in entire proteomes from mass spectrometry data. Following HPLC MS-MS data acquisition, we identify E. coli and S. cerevisiae peptides whose mass and fragment ion spectrum are consistent with that of a peptide bearing a single amino acid substitution, and verify that such spectrum cannot result from a post-translational modification. Our analyses confirm that most substitutions occur due to codon-to-anticodon mispairing within the ribosome. Patterns of errors due to mispairing were similar in bacteria and yeast, suggesting that the error spectrum is chemically constrained. Treating E. coli cells with a drug known to affect ribosomal proofreading increased the error rates due to mispairing at the wobble codon position. Starving bacteria for serine resulted in specific patterns of substitutions reflecting the amino acid deficiency. Overall, translation errors tend to occur at positions that are less evolutionarily conserved, and that minimally affect protein energetic stability, indicating that they are selected against. Genome wide ribosome density data suggest that errors occur at sites where ribosome velocity is relatively high, supporting the notion of a trade-off between speed and accuracy as predicted by proofreading theories. Together our results reveal a mechanistic basis for ribosome errors in translation.

systems biology↗