Gene expression studies rely on high throughput techniques, which do not take in account conceptual limits. I will overcome this situation by exploiting two biological facts. First, RNAs that are important in tissue function are a subset of the global mass, but are always associated with the ribosomal machinery and as such should be identified. Second, gene expression is the outcome of dynamic fluctuations that with time create a unique expression pattern. We need to dynamically label cell populations that undergo stress and follow them to generate a gene expression signature. To achieve my goal, I will consider: 1. Translational stress generated by viral infection or accumulation of misfolded proteins; 2. human CD4+ T lymphocyte subsets which are key to orchestrate immune responses; 3. EIF6 model of metabolic reprogramming.
1. Activation of eIF2alpha phosphorylation by viral infection generates a translational response in which silent mRNAs containing upstream ORFs (uORF) are translated. I will exploit this observation to construct the first in vivo reporter model of translational stress. We will label genetically cells that have translational stress, to identify all the changes that a single cell undergoes after viral infection/accumulation of undegraded proteins.
2. I will selectively sequence for the first time mRNAs and ncRNAs associated with the ribosomal machinery in human cells with a defined functional status.
3. Spectacular data have shown that translation factor eIF6 regulates tumorigenesis by inducing a profound metabolic reprogramming. This observation suggests that, in vivo, translation acts upstream of transcription. We will model how a short translational input results in a complex epigenetic change.
Significance: a revolution in finding biomarkers/drug targets. Generate a map of predictors of the process from stress to disease. Dscriminate biologically active sequences from background. Define how transient translation reshapes gene expression.