Translation decodes genetic information contained in mRNA into functional protein molecules. Despite its importance, the quantitative and mechanistic analysis of translation lags behind other steps of gene expression like transcription, both on the genome-scale and single molecule levels. To fill this gap, our objectives comprise two synergistic pillars: in (2.5.1) we develop next-generation integrated single-molecule and computational tools for monitoring of translation, and in (2.5.2) we leverage those approaches to establish the context-dependency of translation regulatory mechanisms in a physiologically relevant nutrient stress paradigm.
The consortium brings together a unique palette of complementary skills in fluorescent live imaging of RNA and protein molecules (Jeffrey Chao, FMI), systems biology of gene regulation (Felix Naef, EPFL), signal processing and machine learning (Pierre Vandergheynst, EPFL). How the major steps in the translation cycle collectively determine global and gene-specific protein synthesis is highly complex and not well understood in mammalian cells. Moreover, heterogeneity on multiples scales (cells, genes, molecules) and dynamics (stalling), an emerging hallmark of mRNA translation, require combining approaches from omics, live imaging of single molecules, and integration of advanced computational and machine learning approaches.
Subproject 2.5.1 Molecular and computational tools for bulk and single-molecule translation experiments builds a next generation toolbox extending capabilities of ribosome profiling to infer ribosome dwell times using mathematical models (GLMs, TASEP) as a conceptual basis to identify particular translation events (stalling, activation of quality control checkpoints). Moreover, we will use a recent approach from natural language processing (Transformers) to build rich contextual representations of ribosome densities to predict context-dependent translation parameters on an unprecedented scale. Next-generation single-molecule tools for translation site imaging (smGCN4, TREAT-smGCN4, smGCN4+HA-Fb) will enable measurements of initiation and elongation dynamics on individual mRNAs, including simultaneous assessment of translation and mRNA decay. Accompanying signal processing methods for optimal detection and time-series modeling of individual translation sites will allow to quantify translation heterogeneity.
Subproject 2.5.2 Quantitative dissection of the translation cycle in response to nutrient stress leverages these tools to address how global vs. gene-specific translation is rewired under perturbed nutrient availability. First, we dissect responses of the translation cycle to systematic deprivation of essential amino acids using calibrated ribosome profiling, and apply the computational models to detect altered global or gene-specific translation events (ribosome stalling and traffic, QC checkpoints). Single-molecule imaging follow-ups, and in particular links between translation and mRNA stability will be dissected with smGCN4 and TREAT. Finally, a massively parallel two-color cell-based translation reporter assay will provide an orthogonal approach to identify determinants and importantly downstream effects of ribosome stalling on protein levels.
Last updated:22.03.2023
