Systematic analysis of gene expression at the level of single RNA molecules – HI-FISH

Gene expression is of fundamental importance in all aspects of a cell’s life, and describing it at the highest level of precision and sensitivity is a major focus of the scientific community. This has led to key technological developments such as microarrays or massive parallel sequencing. With these tools, it has been possible to: (i) discover many new RNA families; (ii) perform expression profiling in normal and pathological cells; and (iii) study gene regulation genome-wide, thereby revolutionizing systems biology. Despite their power, these techniques suffer from several limitations: (i) cells are lysed and spatial information of RNA localization is lost; (ii) single cell studies are difficult, because few cells can be analyzed in parallel (typically up to 100), and low expressing genes are nearly impossible to quantify. Single molecule FISH (smFISH) identifies and localizes all mRNA molecules produced by a given gene, in every cell of large populations. It thus possesses unique advantages to study gene expression. In this project, we will use and develop smFISH to investigate expression of thousands of human genes in native fixed cells.

RNA localization: a fundamental but understudied phenomenon
Most mRNAs distribute throughout the cytoplasm in a random manner, but some localize to specific subcellular areas. Specific localizations have been observed from bacteria to humans, and they occur in cells as diverse as oocytes/early embryos, neurons and other somatic cells. RNA localization is linked to either RNA metabolism, for instance when mRNAs are stored in P-bodies; or to protein metabolism, to synthesize a protein locally. Such localizations are involved in: (i) cell polarity, including cell migration, asymmetric cell division, and localized signaling; (ii) co-translational assembly of protein complexes; (iii) protein targeting to cellular organelles; (iv) synaptic plasticity, etc …
Although dozens of localized mRNAs have been identified across species, there are only two systematic FISH studies, including one performed in Drosophila embryos where 3000 genes have been studied. This showed that nearly 60% of the mRNAs studied displayed atypical, non-random subcellular localization patterns. This indicates that many important discoveries remain to be made, in particular in vertebrates where information is scarce: new localized mRNAs, new localization patterns, and novel functions for this process.

Transcriptional noise: an important determinant of cellular phenotype.
Even in clonal populations, there are differences in gene expression between individual cells. These variations can be a response to a particular microenvironment, but they can also arise from spontaneous, stochastic fluctuations of promoter activity, often referred to as transcriptional noise. Noise in gene expression contributes to the phenotypic variability of clonal cells and multicellular organisms. There are however only a handful of studies that used smFISH to measure it, and only few genes have been analyzed.

Research plan
Our goal is to use smFISH at a large scale to explore the human genome. First, we will use a library of cell lines that each expresses a tagged mRNA. We will use probes against the tag to perform smFISH for all genes in this library, and we will develop image analysis tools specifically dedicated to high-throughput smFISH. This will allow us to identify mRNAs having atypical, non-random cytoplasmic localization, and we will conduct functional studies on some of these genes. We will also develop new smFISH techniques, in order to make them compatible with the high-throughput detection of untagged, endogenous mRNAs. These technological breakthroughs will enable us to reach additional objectives, such as global analysis of transcriptional noise and in the mid-term, the analysis of the entire human genome. High-throughput smFISH will open new research directions and will provide a systems view of RNA localization and expression noise.