From transcriptional noise to the sound of regulation – CutReg

Hidden transcription is widespread in the genomes of eukaryotes, but the mechanisms from which it originates and its functional significance are still unclear. Cryptic unstable transcripts (CUTs) are a new class of RNA Pol II transcripts that we have recently identified (Wyers et al., 2005) and that might represent the largest share of hidden transcription in the genome of S.cerevisiae. These non-coding RNAs are transcribed ubiquitously and are degraded rapidly and efficiently in the nucleus. Whether CUTs are the mere result of spurious transcriptional activity or the product of a pathway with functional significance is poorly understood. What is the share of significant information that is encrypted in CUTs is also unclear to date. The aim of this project is to study the metabolism and the function of CUTs. We propose to tackle several major questions raised by the existence of these genetic elements utilizing multidisciplinary approaches ranging from genome-wide analyses to genetics and biochemistry. Starting from the virtually complete genome-wide distribution of CUTs in exponentially growing cells obtained in one of our laboratories (Neil et al., in preparation) we will perform bioinformatic analyses to correlate the occurrence of hidden transcription (i.e. transcription without the production of stable RNAs) to other published genomic features, including nucleosomes distribution, the occurrence of chromatin modification marks and the position of bona fide genes. The interest of these studies will be double: on one side they might validate on statistical bases some of the models proposed to explain CUTs "raison d'être"; second, and most importantly, they will provide essential clues to formulate new hypotheses that will be validated with targeted studies on chosen model cases. The results obtained from these specific cases will then be precious for a feed-back refinement of the genomewide analyses. A second facet will concern the metabolism of CUTs, i.e. the mechanisms of their production and their degradation. Although we have progressed significantly in the understanding how these RNAs are recognized and targeted for degradation (Thiebaut et al., 2006), several important questions remain unanswered. The 'stop-and-degrade' information contained in the known signals within CUTs sequences is unlikely sufficient to ensure an efficient and error-free mechanism: we need more information on the function of the termination and degradation system that takes into account the chromatin environment because our genome-wide analysis suggests an implication of nucleosomes in this process. We propose genetic and biochemical approaches to tackle these questions. Why are CUTs transcribed in the first place is also unclear: are there specific and regulated promoters? Are they byproducts of bi-directional promoters designed to transcribe divergent genes or of spurious transcription randomly arising in nucleosome-free regions? We will benefit from genome-wide studies, single-gene analyses and artificially designed system to answer these questions. An important facet of this proposal is the study of the function of CUTs. A fascinating hypothesis is that an inherently imprecise mechanism (i.e. spurious transcription initiation) is both controlled a posteriori by the termination/degradation machinery and diverted to serve regulatory functions in the cells, spanning from the control of gene expression to the maintenance of genomic stability (Houseley et al., 2007; Vasiljeva et al., 2008). We have identified a novel example of regulation that implicates CUTs production to modulate the expression of genes involved in the nucleotide biogenesis pathway (Thiebaut et al., submitted). The study of the regulation of the URA2 and IMD2 model systems will be pursued in the aim of resolving the many questions that remain open. Important implications are expected from these studies that are likely to impact and modify the current models on the mechanism and regulation of transcription initiation. Our genome-wide analyses suggest that this mode of regulation extends to other regulons, potentially implicating a wider impact of CUTs on the expression of the transcriptome. We will undertake a large-scale analysis of CUTs expression under different metabolic conditions, notably in stationary phase and upon exit from stationary phase. One of our laboratories has a long-standing experience in the study of gene expression under stationary phase and has particularly characterized the expression of the IMD2 gene in these conditions. We will join efforts to understand regulation during stationary phase in the perspective of regulatory transcription and CUTs production. We believe that our teams possess the human potential and the technical know-how to ensure the necessary feedback between large-scale analyses, small-scale validation and the mechanistic analyses that constitute the focal point of this proposal.