Epitranscriptomic analyses of the dihydrouridine RNA modification in Bacteria and Eukarya

Promoter

Dr. Damien HERMAND, UNamur, Molecular Physiology Research Unit (URPhyM), Molecular Genetics (GeMo)

Jury

Damien Hermand (Supervisor), Olivier De Backer (President), Nicolas Gillet (Jury), Marc Graille (Jury), Régis Hallez (Jury) & Denis Lafontaine (Jury)

Summary

A large catalogue of modified nucleosides is found within RNAs. Most of these chemical modifications have been known for the past fifty years. However, the first epitranscriptomic studies are recent and shed light to a new layer of gene expression regulation. To date, less than ten RNA modifications have been studied at the transcriptomic level due to the difficulty to determine their distribution at the single- nucleotide resolution.
The dihydrouridine RNA modification (D) is a product of the reduction of uridine by a conserved family of dihydrouridine synthases (Dus) and has been exclusively studied within the context of tRNA and rRNA biology. Despite its universal conservation among the three domains of life, little is known about the biological relevance of this modification.
Taking advantage of the previously reported fluorescent labeling of tRNAs at dihydrouridylated positions, we have developed D-seq (dihydrouridine-sequencing), a sensitive method to detect the transcriptome-wide distribution of dihydrouridines, which we confirmed with dot blot and primer extension assays. Remarkably, the D-seq pipeline could be easily applied to other RNA modifications that are sensitive to the rhodamine labeling and for which the cognate RNA modifying enzyme mutants are available.
Among the 372 putative dihydrouridylated sites found on the yeast transcriptome, 60% were located on well-known positions of tRNAs. Surprisingly, the remaining predicted sites were found onto messenger RNAs. The dissection of D distribution revealed an enrichment on coding sequences and a non-random distribution on codons. Unlike coding RNAs, hypodihydrouridylated tRNAs were remarkably more abundant than their modified counterparts. The conservation of this feature in another eukaryote is currently under investigation on a human transcriptome.
Interestingly, the E. coli coding transcriptome was devoid of any detectable dihydrouridine in the tested condition, implying that mRNA dihydrouridylation could be eukaryotic-specific. Our data also suggest that a novel yet uncharacterized bacterial Dus enzyme modifies the 23S rRNA and might target other RNA species.
Altogether, our data demonstrate the reliability of the unbiased D-seq technique to map the dihydrouridine at the single-nucleotide resolution. Moreover, we propose dihydrouridine as a potential new internal mRNA modification in Eukarya.