ERC FUNDED PROJECTS

The metabolism of cancer cells is altered to meet cellular requirements for growth, providing novel means to selectively target tumorigenesis. While extensively studied, our current view of cancer cellular metabolism is fundamentally limited by lack of information on variability in metabolic activity between distinct subcellular compartments and cells. We propose to develop a spatio-temporal fluxomics approach for quantifying metabolic fluxes in the cytoplasm vs. mitochondria as well as their cell-cycle dynamics, combining mass-spectrometry based isotope tracing with cell synchronization, rapid cellular fractionation, and computational metabolic network modelling. Spatio-temporal fluxomics will be used to revisit and challenge our current understanding of central metabolism and its induced adaptation to oncogenic events – an important endeavour considering that mitochondrial bioenergetics and biosynthesis are required for tumorigenesis and accumulating evidences for metabolic alterations throughout the cell-cycle. Our preliminary results show intriguing oscillations between oxidative and reductive TCA cycle flux throughout the cell-cycle. We will explore the extent to which cells adapt their metabolism to fulfil the changing energetic and anabolic demands throughout the cell-cycle, how metabolic oscillations are regulated, and their benefit to cells in terms of thermodynamic efficiency. Spatial flux analysis will be instrumental for investigating glutaminolysis - a ‘hallmark’ metabolic adaptation in cancer involving shuttling of metabolic intermediates and cofactors between mitochondria and cytoplasm. On a clinical front, our spatio-temporal fluxomics analysis will enable to disentangle oncogene-induced flux alterations, having an important tumorigenic role, from artefacts originating from population averaging. A comprehensive view of how cells adapt their metabolism due to oncogenic mutations will reveal novel targets for anti-cancer drugs.

1 481 250 €

Start date: 2017-02-01, End date: 2022-01-31

Over the past decade small RNAs such as microRNAs (miRNAs) and small interfering RNAs (siRNAs) have been shown to carry pivotal roles in post-transcriptional regulation and genome protection and to play an important part in various physiological processes in animals. miRNAs can be found in a very wide range of animals yet their functions were studied almost exclusively in members of the Bilateria such as insects, nematodes and vertebrates. Hence studying their function in representatives of non-bilaterian phyla such as Cnidaria (sea anemones, corals, hydras and jellyfish) is crucial for understanding the evolution of miRNAs in animals and can provide important insights into their roles in the ancient ancestor of Cnidaria and Bilateria. The sea anemone Nematostella vectensis is an excellent model for such a study since it can be grown in large numbers throughout its life cycle in the lab and because well-established genetic manipulation techniques are available for this species. Our preliminary results indicate that miRNAs in Nematostella frequently have a nearly perfect match to their messenger RNA (mRNA) targets, resulting in cleavage of the target. This mode of action is common for plant miRNAs, but is very rare in Bilateria. This finding together with my recent discovery of a Nematostella homolog of HYL1, a protein involved in miRNA biogenesis in plants, raises the exciting possibility that the miRNA pathway existed in the common ancestor of plants and animals. Here I suggest to bring together an array of advanced biochemical and genetic methods such as gene knockdown, transgenesis, high throughput sequencing and immunoprecipitation in order to obtain - for the first time - a deep understanding of the biogenesis and mechanism of action of small RNAs in Cnidaria. This will provide a novel way to understand the evolution of this important molecular pathway and to evaluate its age and ancestral form.

1 499 587 €

Start date: 2015-05-01, End date: 2020-04-30

Transcription factors (TF) regulate genome function by controlling gene expression. Comprehensive characterization of the in vivo binding of TF to the DNA in relevant primary models is a critical step towards a global understanding of the human genome. Recent advances in high-throughput genomic technologies provide an extraordinary opportunity to develop and apply systematic approaches to learn the underline principles and mechanisms of mammalian transcriptional networks. The premise of this proposal is that a tractable set of rules govern how cells commit to a specific cell type or respond to the environment, and that these rules are coded in regulatory elements in the genome. Currently our understanding of the mammalian regulatory code is hampered by the difficulty of directly measuring in vivo binding of large numbers of TFs to DNA across multiple primary cell types and their natural response to physiological stimuli. Here, we overcome this bottleneck by systematically exploring the genomic binding network of 1. All relevant TFs of key hematopoietic cells in both steady state and under relevant stimuli. 2. Follow the changes in TF networks as cells differentiate 3. Use these models to engineer cell states and responses. To achieve these goals, we developed a new method for automated high throughput ChIP coupled to sequencing (HT-ChIP-Seq). We used this method to measure binding of 40 TFs in 4 time points following stimulation of dendritic cells with pathogen components. We find that TFs vary substantially in their binding dynamics, genomic localization, number of binding events, and degree of interaction with other TFs. The analysis of this data suggests that the TF network is hierarchically organized, and composed of different types of TFs, cell differentiation factors, factors that prime for gene induction, and factors that bind more specifically and dynamically. This proposal revisits and challenges the current understanding of the mammalian regulatory code.

1 500 000 €

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

Computational, structure-based, drug design offers insight at an atomic resolution, which is commonly not attainable by experimental means. Detailed calculations on protein-ligand interactions help to rationalize and predict experimental findings. Accurate and efficient calculations of binding free energies is essential in this respect. In addition, knowledge concerning the enthalpic and entropic contributions are highly relevant to determine novel drug design strategies and to understand the underlying principles of ligand binding. Currently available methods to address ligand affinity either do not include all relevant contributions to the binding free energy, or are too computationally demanding to be applied straightforwardly. In addition, calculations on enthalpy and entropy for drug design purposes are very rare, due to the difficulty in calculating these accurately. This proposal describes the research that leads the way to new, standard applications to be used in drug design processes in academia and industry. Furthermore, we propose to investigate the enthalpic and entropic contributions to ligand binding. We define a ligand-surroundings enthalpy and entropy, which conveys more information than the experimentally accessible enthalpy and entropy of ligand binding. In support of this research, we will develop new enhanced sampling techniques which not only render the above calculations practically feasible, but which will also find their application in related research questions such as the protein folding problem or the elucidation of protein-protein interactions. The methods described are highly relevant for the pharmaceutical industry, where currently available computational approaches are insufficient to answer the questions of todays drug discovery programmes.

1 485 615 €

Start date: 2011-01-01, End date: 2015-12-31

"Our existence as human beings is based on plants and their products. Worldwide, crops are threatened by pests including biotrophic fungi. Therefore, it is of vital interest to develop new strategies to reduce crop losses and to improve crop plants for the growing world population. Biotrophic plant pathogens employ small secreted molecules, so-called effectors, to overcome plant defence systems and to establish biotrophy. The rapid increase in available genome sequences of biotrophic pathogens and in transcriptomic datasets of their biotrophic stages allow us to identify putative secreted proteinaceous effectors by bioinformatic means. However, our insight into the functions of these effectors is still very limited. In this proposal, the PI´s extensive experience on both the plant host side and the fungal pathogen side of the biotrophic interaction is exploited to develop a workflow for functional, partially robotic-based screens to fill this gap. The combination of screen-deduced functional information with the analysis of effector localisation and specific host interactors will provide the basis for formulating starting hypotheses of effector function. These will then be tested in individual case studies, employing the well established Ustilago maydis-Zea mays as well as the new Ustilago bromivora-Brachypodium distachyon model systems. The project will be conducted at the Max Planck Institute (MPI) for Terrestrial Microbiology in a highly stimulating scientific environment. Linking the dramatic morphological changes and underlying molecular events during biotrophy on the host side to the action of subsets or even single effector proteins will allow the creation of a synthetic effectome. The deep functional understanding of the manipulative toolbox of biotrophs has the potential to facilitate transgenic crop development and will open a new era in the development of sustainable antifungal plant protection strategies."

1 446 316 €

Start date: 2014-02-01, End date: 2019-01-31

Animals constantly face complex environments consisted of multiple fluctuating cues. Accurate detection and efficient integration of such perplexing information are essential as animals’ fitness and consequently survival depend on making the right behavioral decisions. However, little is known about how multifaceted stimuli are integrated by neural systems, and how this information flows in the neural network in a single-neuron resolution. Here we aim to address these fundamental questions using C. elegans worms as a model system. With a compact and fully-mapped neural network, C. elegans offers a unique opportunity of generating novel breakthroughs and significantly advance the field. To study functional dynamics on a network-wide scale with an unprecedented single-neuron resolution, we will construct a comprehensive library of transgenic animals expressing state-of-the-art optogenetic tools and Calcium indicators in individual neurons. Moreover, we will study the entire encoding process, beginning with the sensory layer, through integration in the neural network, to behavioral outputs. At the sensory level, we aim to reveal how small sensory systems efficiently encode the complex external world. Next, we will decipher the design principles by which neural circuits integrate and process information. The optogenetic transgenic animals will allow us interrogating computational roles of various circuits by manipulating individual neurons in the network. At the end, we will integrate the gathered knowledge to study how encoding eventually translates to decision making behavioral outputs. Throughout this project, we will use a combination of cutting-edge experimental techniques coupled with extensive computational analyses, modelling and theory. The aims of this interdisciplinary project together with the system-level approaches put it in the front line of research in the Systems Biology field.

1 498 400 €

Start date: 2014-02-01, End date: 2019-01-31

Epigenetic alterations can be detected in all cancers and in essentially every patient. Despite their prevalence, the concrete functional roles of these alterations are not well understood, for two reasons: First, cancer samples tend to carry many correlated epigenetic alterations, making it difficult to statistically distinguish relevant driver events from those that co-occur for other reasons. Second, we lack tools for targeted epigenome editing that could be used to validate biological function in perturbation and rescue experiments. The proposed project strives to overcome these limitations through experimental and bioinformatic methods development, with the ambition of making and breaking cancer cells in vitro by introducing defined sets of epigenetic alterations. We will focus on leukemia as our “model cancer” (given its low mutation rate, frequent defects in epigenetic regulators, and availability of excellent functional assays), but the concepts and methods are general. In Aim 1, we will generate epigenome profiles for a human knockout cell collection comprising 100 epigenetic regulators and use the data to functionally annotate thousands of epigenetic alterations observed in large cancer datasets. In Aim 2, we will develop an experimental toolbox for epigenome programming using epigenetic drugs, CRISPR-assisted recruitment of epigenetic modifiers for locus-specific editing, and cell-derived guide RNA libraries for epigenome copying. Finally, in Aim 3 we will explore epigenome programming (methods from Aim 2) of candidate driver events (predictions from Aim 1) with the ultimate goal of converting cancer cells into non-cancer cells and vice versa. In summary, this project will establish a broadly applicable methodology and toolbox for dissecting the functional roles of epigenetic alterations in cancer. Moreover, successful creation of a cancer that is driven purely by epigenetic alterations could challenge our understanding of cancer as a genetic disease.

1 281 205 €

Start date: 2016-12-01, End date: 2021-11-30

"Gene expression in living cells is a most intricate molecular process, occurring in stages, each of which is regulated by a diversity of mechanisms. Among the various stages leading to gene expression, only transcription is relatively well understood, thanks to Genomics and bioinformatics. In contrast to the vast amounts of genome-wide data and a growing understanding of the structure of networks controlling transcription, we still lack quantitative, genome-wide knowledge of the mechanisms underlying regulation of mRNA degradation and translation. Among the unknowns are the identity of the regulators, their kinetic modes of action, and their means of interaction with the sequence features that make-up their targets; how these target combine to produce a higher level ""grammar"" is also unknown. An important part of the project is dedicated to generating genome-wide experimental data that will form the basis for quantitative and more comprehensive analysis of gene expression. Specifically, the primary objectives of our proposed research plan are: 1) to advance our understanding of the transcriptome, by deciphering the code regulating mRNA decay 2) to break the code which controls protein translation efficiency 3) to understand how mRNA degradation and translation efficiency determine noise in protein expression levels. The proposed strategy is based on an innovative combination of computational prediction, synthetic gene design, and genome-wide data acquisition, all culminating in extensive data analysis, mathematical modeling and focused experiments. This highly challenging, multidisciplinary project is likely to greatly enhance our knowledge of the various modes by which organisms regulate expression of their genomes, how these regulatory mechanisms are interrelated, how they generate precise response to environmental challenges and how they have evolved over time."

1 320 000 €

Start date: 2008-09-01, End date: 2013-08-31

Our working hypothesis is that tumorigenesis is an evolutionary process that fundamentally couples few major driving events (point mutations, rearrangements) with a complex flux of minor aberrations, many of which are epigenetic. We believe that these minor events are critical factors in the emergence of the cancer phenotype, and that understanding them is essential to the characterization of the disease. In particular, we hypothesize that a quantitative and principled evolutionary model for carcinogenesis is imperative for understanding the heterogeneity within tumor cell populations and predicting the effects of cancer therapies. We will therefore develop an interdisciplinary scheme that combines theoretical models of cancer evolution with in vitro evolutionary experiments and new methods for assaying the population heterogeneity of epigenomic organization. By developing techniques to interrogate DNA methylation and its interaction with other key epigenetic marks at the single-cell level, we will allow quantitative theoretical predictions to be scrutinized and refined. By combining models describing epigenetic aberrations with direct measurements of chromatin organization using Hi-C and 4C-seq, we shall revisit fundamental questions on the causative nature of epigenetic changes during carcinogenesis. Ultimately, we will apply both theoretical and experimental methodologies to assay and characterize the evolutionary histories of tumor cell populations from multiple mouse models and clinical patient samples.

1 499 998 €

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