Conformational dynamics in microRNAs : the example of miR-34a targeting Sirt1 mRNA

Abstract: In biology, regulatory mechanisms are essential to achieve complex tasks, as virtually every process can be positively or negatively modulated in its outcome, upon different cues. In humans, microRNAs (miRNAs) constitute a fundamental layer of post-transcriptional gene expression regulation. This class of molecules finely tune protein expression, by downregulating messenger RNAs (mRNA) levels and their translation. The mechanism by which miRNAs find and act upon their targets primarily relies on their nucleotide sequence, relative to the corresponding binding site on the mRNA. The development of an exhaustive miRNA–mRNA interactome is particularly attractive because of the profound implication for basic biology as well as for diagnostics and therapeutics in human health. However, computational prediction of target sites and associated downregulation levels, using the limited sequence determinants available, is still an outstanding challenge in the field. In this thesis, we bring forward the hypothesis that modeling of miRNA–mRNA pairs might benefit from considering the inherent structural flexibility of these complexes, at the molecular level. In the introductory chapter, we present the structural features of RNAs with a focus on their conformational dynamics and NMR spectroscopy as a tool to investigate these motions. The molecular details of miRNA biogenesis and function are later introduced to contextualize the results of Paper I. Finally, the challenges associated with RNA sample preparation are discussed in light of the work presented in Paper II. In Paper I, we show that a miRNA–mRNA pair involved in a cancer-regulating pathway exploits its flexibility to toggle between lower and higher target repression states. This study shows that suboptimal structures of a given miRNA–mRNA pair, that are overlooked by computational prediction and that often elude experimental detection, can be functionally relevant and are essential to draw a mechanistic picture of miRNA function. The methods used in Paper I for RNA sample preparation and molecular simulation are described in Paper II and II, respectively. While these methods were essential to achieve the results of Paper I, they also find widespread application in the RNA field.