ERC FUNDED PROJECTS

A major aim in genome research is to reveal how genetic variation affects phenotypic variation. Here I propose to use high-throughput genomics (whole genome sequencing, transcriptome and epigenome analysis) to screen carefully selected study populations where the chances are particularly favourable to obtain novel insight into genotype-phenotype relationships. The ambition is to take discoveries all the way from phenotypic characterization to the identification of the genes and the actual genetic variant causing a phenotypic effect and to understanding the underlying functional mechanisms. The program will involve a fish (the Atlantic herring), a bird (the domestic chicken) and a mammal (the European rabbit). The Atlantic herring will be studied because it provides unique opportunities to study the genetics of adaptation in a natural population and because of the possibilities to revolutionize the fishery management of this economically important marine fish. We will generate a draft assembly of the herring genome and then perform whole genome resequencing of different populations to reveal the population structure and the loci underlying genetic adaptation. The European rabbit is an excellent model for studying the genetics of speciation due to the presence of two distinct subspecies on the Iberian Peninsula. The domestication of the rabbit is also particularly interesting because it is a recent event (about 1500 years ago) and it is well established that domestication happened from the wild rabbit population in southern France. Finally, the domestic chicken provides excellent opportunities for in depth functional studies since it is both a domestic animal harbouring a rich genetic diversity and an experimental organism. (BATESON is the acronym for this proposal because Bateson (1902) pioneered the study of genotype-phenotype relationships in animals and used the chicken for this work.)

2 300 000 €

Start date: 2012-05-01, End date: 2017-04-30

Embryogenesis is the temporal unfolding of cellular processes: proliferation, migration, differentiation, morphogenesis, apoptosis and functional specialization. These processes are well understood in specific tissues, and for specific cell types. Nevertheless, our systematic knowledge of the types of cells present in the developing and adult animal, and about their functional and lineage relationships, is limited. For example, there is no consensus on the number of cell types, and many important stem cells and progenitors remain to be discovered. Similarly, the lineage relationships between specific cell types are often poorly characterized. This is particularly true for the mammalian nervous system. We have developed (1) a reliable high-throghput method for sequencing all transcripts in 96 single cells at a time; and (2) a system for high-throughput phylogenetic lineage reconstruction. We now propose to characterize embryogenesis using a shotgun approach borrowed from genomics. Tissues will be dissected from multiple stages and dissociated to single cells. A total of 10,000 cells will be analyzed by RNA sequencing, revealing their functional cell type, their lineage relationships, and their current state (e.g. cell cycle phase). The novel approach proposed here will bring the powerful strategies pioneered in genomics into the field of developmental biology, including automation, digitization, and the random shotgun method. The data thus obtained will bring clarity to the concept of ‘cell type’; will provide a first catalog of mouse brain cell types with deep functional annotation; will provide markers for every cell type, including stem cells; and will serve as a basis for future comparative work, especially with human embryos.

1 496 032 €

Start date: 2010-11-01, End date: 2015-10-31

The aim of this project is to understand the dynamics of protein-DNA interactomes underlying circadian oscillators in mammals, and how these shape circadian transcriptional output programs. Specifically our goal is to solve a fundamental issue in circadian biology: the phase specificity problem underlying circadian gene expression. We have taken a challenging and original multi-disciplinary approach in which molecular biology experiments will be tightly interlinked with computational analyses and biophysical modeling. The approach will generate time resolved protein-DNA interactomes in mouse liver for several key circadian repressors at unprecedented resolution. These experiments will be complemented with chromosome conformation capture (3C) experiments to monitor how looping interactions and 3D genome structure rearrange during the circadian cycle, which will inform on how circadian transcription networks use long-range gene regulatory mechanisms. Novel computational algorithms based on biophysical principles will be developed and implemented to optimally analyze interactome and 3C datasets. For the latter, statistical models from polymer physics will be used to reconstruct the chromatin networks and interaction maps from the 3C data. At the detailed level of individual cells, we will investigate transcription bursts, and how those are involved in the control of circadian gene expression. In particular we will exploit high temporal resolution bioluminescence reporters using a biophysical model of transcription coupled with a Hidden Markov Model (HMM). Through our innovative approach, we expect that the data generated and state-of-the-art analyses performed will lead novel insight into the role and mechanics of circadian transcription in controlling circadian outputs in mammals.

1 500 000 €

Start date: 2011-03-01, End date: 2016-02-29

Differentiation events in mammalian development involve stable resetting of transcriptional programs, which entails changes in the epigenetic state of target sequences defined by modifications of DNA and bound nucleosomes. These recently identified epigenetic layers modulate DNA accessibility in a positive and negative manner and thus could make genetic readouts context-dependent and dynamic. The proposed project aims to quantify the epigenetic contribution to cellular differentiation as a key event in development. By applying parallel genomic approaches we will comprehensively define the epigenome and its plasticity during cellular commitment of pluripotent murine stem cells into defined terminally differentiated cells. We will focus on DNA methylation and its interplay with several histone modifications as a way to achieve stable gene silencing. The resulting global profiles will gain insights into targeting principles and generate statistical, predictive models of regulation. From these mechanistic models will be derived and tested by genetically interfering with genetic and epigenetic regulatory pathways and by dissecting DNA sequence components involved in specifying targets. These experiments aim to unravel the crosstalk between epigenetic regulation and cell plasticity, the underlying molecular circuitry in pluripotent and unipotent cells and ultimately help to incorporate epigenetic regulation into current transcriptional regulatory models.

1 085 000 €

Start date: 2008-10-01, End date: 2013-09-30

Two of greatest challenges of the post-genomic era are to (i) develop a detailed understanding of the heritable variation in the human genome, and to (ii) determine which key events in human evolutionary history that are responsible for patterns of genomic variation. The recent genomic revolution will be instrumental in these quests and we will very soon have access to several thousand complete genomes from diverse populations. Extracting genetic information from ancient material has for long been hampered by numerous difficulties after its first steps some two decades ago, but in the last few years, many of these problems have been solved and the use of ancient DNA is now beginning to show its full potential. The use of ancient DNA has been announced among the top-ten ‘insights of the decade’ by Science, and promises to transform our views on human origins and prehistory. The demographic history of Europeans attracts great interest in archaeology, anthropology, and human genetics, and it has drawn extensive research focus for more than a century. The recent genomic revolution has opened up the time dimension for genomic analyses, however, to harness the full potential of genomic data from modern and ancient material, we need new population genetic theory and modern statistical analysis tools. I propose to conduct 3 Ancient Genome Projects to generate complete genomes for multiple individuals from 3 time epochs in the European prehistory; the Cro-Magnon-, the Mesolithic-, and the Neolithic-Genome project. These Genome Projects will proceed in concert with development a) new population genetic theory and novel tools for demographic inference, b) a novel, temporal based, framework for characterizing selection and local adaptation, and c) explore the evolutionary history of gene-variants associated with traits and diseases. Genomic data from temporal samples has the potential to revolutionize our understanding of human evolution and the demographic history of Europe.

1 500 000 €

Start date: 2012-10-01, End date: 2017-09-30

In human cells, two meters of DNA sequence are compressed into a nucleus whose linear size is five orders of magnitude smaller. Deciphering how this amazing structural organization is achieved and how DNA functions can ensue in the environment of a cell’s nucleus represent central questions for contemporary biology. Here, I embrace this challenge by establishing a comprehensive framework of microscopy and sequencing technologies coupled with advanced analytical approaches, aimed at addressing three fundamental highly-interconnected questions: 1) What are the design principles that govern DNA compaction? 2) How does genome structure vary between different cell types as well as among cells of the same type? 3) What is the link between genome structure and function? In preliminary experiments, we have devised a powerful method for Genomic loci Positioning by Sequencing (GPSeq) in fixed cells with optimally preserved nuclear morphology. In parallel, we are developing high-end microscopy tools for simultaneous localization of dozens of genomic locations at high resolution in thousands of single cells. We will obtain first-ever genome-wide maps of radial positioning of DNA loci in the nucleus, and combine them with available DNA contact probability maps in order to build 3D models of the human genome structure in different cell types. Using microscopy, we will visualize chromosomal shapes at unprecedented resolution, and use these rich datasets to discover general DNA folding principles. Finally, by combining high-resolution chromosome visualization with gene expression profiling in single cells, we will explore the link between DNA structure and function. Our study shall illuminate the design principles that dictate how genetic information is packed and read in the human nucleus, while providing a comprehensive repertoire of tools for studying genome organization.

1 499 808 €

Start date: 2018-01-01, End date: 2022-12-31

The main scientific questions addressed in this proposal relate to the understanding of molecular mechanisms of growth control and cancer through the combined use of high-throughput technologies and computational biology. We aim to create a systems-level understanding of the cell cycle, and its regulation by physiological growth factors and oncogenes through the use high-throughput biology to identify all or the majority of genes that are essential for cell cycle progression, and by combining this dataset with computationally predicted and experimentally validated target genes of growth factors and oncogenic pathways. In my opinion, such systems biology approach is critical for understanding of growth control, as organ-specific growth control has proven particularly refractory to genetic dissection. Much of what we know about physiological mechanisms controlling cellular growth in mammals has been revealed by human cancer genetics. These studies have revealed that a large number of genes can contribute to aberrant cell growth; there are more than 300 genes that have been linked to cancer, and mutations found in cancer are often cell type specific ( oncogene preference , i.e. PTCH mutations in medulloblastoma, APC in colon cancer, TMPRSS2-ERG in prostate cancer), suggesting that different pathways in different cell lineages are coupled to the cell cycle machinery. We have preliminary evidence that hedgehog (Hh) and Wnt signals are directly coupled to expression of N-myc and c-Myc genes, but only in tissues and cell-types that display a proliferative response to these factors. Both classical molecular and developmental biology as well as high throughput and systems biological methods will be used for dissection of the molecular mechanism of this selectivity. If successful, these experiments would establish a principle explaining why particular mutations are extremely common in some tumor types but not found at all in others.

2 200 000 €

Start date: 2009-03-01, End date: 2014-02-28

The domestic dog encompasses hundreds of genetically isolated breeds, many of which show an increased risk for certain diseases. With the canine genome sequence, an understanding of the haplotype structure and availability of disease gene mapping tools, we are now in a unique position to map canine disease genes to inform human biology and medicine. So far we have mapped monogenic traits as well as >40 loci for >10 complex traits. We now propose to map genes for key diseases using many breeds to dissect a larger number of genes underlying the specific disease. We further plan to evaluate the functional consequences of mutations and pilot personalized treatment strategies based on genetic risk. The specific aims are: 1.Characterization of disease phenotypes, breed predisposition and sample acquisition. We are currently collecting samples from >20 diseases and will expand our phenotypic classification and sample collection to a larger number of breeds for some key diseases such as osteosarcoma, breast cancer, behavior and atopy or lymphocytic thyroiditis. 2.Identification and functional characterization of canine disease genes and pathways. We will perform genomewide association mapping followed by targeted resequencing for mutation detection. Pathway analysis will be performed to understand the disease mechanisms mostly contributing to the disease. For select mutations, we will use state of the art molecular biology to provide detailed functional characterization of selected genes revealed by our gene discovery platform. 3.Piloting canine personalized treatment strategies based on inherited risk factors. For a few diseases we will pilot personalized treatment strategies based on inherited risk factors, utilizing the genetic information gathered in aim 2. Available or novel drugs acting on the identified pathways will be tested in dogs with specific risk factors using a veterinary network for clinical trials. Knowledge gained should inform human personalized medicine.

1 499 365 €

Start date: 2012-12-01, End date: 2017-11-30

It is now becoming apparent that genes are regulated not only by transcription, but also by thousands of post-transcriptional regulators that can stabilize or degrade mRNAs. Some of the most important regulators are miRNAs, short RNA molecules that are deeply conserved in sequence and are involved in numerous biological processes, including human disease. Surprisingly, transcriptomic and proteomic studies show that most miRNAs only have subtle silencing effects on their targets, suggesting additional important, but yet undiscovered functions. Thus the question is raised: if the main function of miRNAs is not to silence targets, what is it? I will test two novel hypotheses about miRNA function. The first hypothesis proposes that miRNAs can buffer gene expression noise. The second hypothesis is inspired by my preliminary results and proposes that miRNAs can synchronize expression of genes. If I validate either hypothesis, it would mean that miRNA functions can be investigated in entirely new ways, yielding important new biological insights relevant to both basic research and human health. However, these hypotheses can only be tested in individual cells, and the necessary single-cell technologies and computational tools are only maturing now. I will apply my expertise in miRNA biology and in combined wet-lab and computational methods to design, develop and apply miRCell-seq to test these two hypotheses in cell cultures and in animals. This new method will for the first time measure miRNAs, their targets, and the interactions between them in single cells and transcriptome-wide. We will use mutant cells devoid of miRNAs and time course experiments to generate sufficient data to develop detailed models of the miRNA impact on their targets. We will then validate our findings with single cell proteomics. This project thus has the potential to reveal novel functions of miRNAs and substantially improve our general understanding of gene regulation.

1 497 650 €

Start date: 2018-03-01, End date: 2023-02-28