Albertha J. Walhout, Ph.D. - US grants

DESCRIPTION (provided by applicant): Obesity is the fastest growing health problem in the US and is associated with an increased risk of type 2 diabetes, heart disease and certain forms of cancer. Obesity is, in part, a genetic disorder and several genes that are involved in the disease have been identified. However, these likely constitute only the "tip of the iceberg" because more potential fat genes are being identified in model organisms. For example, approximately 400 novel genes involved in fat metabolism (hereafter referred to as "fat genes") have recently been identified in the nematode Caenorhabditis elegans. One of the main questions is how these genes function together in fat metabolism and whether any of these genes may be suitable therapeutic targets for the treatment or prevention of obesity. In both humans and nematodes, several fat genes encode regulatory transcription factors ("fat TFs"). We hypothesize that some of these TFs regulate the expression of many fat genes and that their human homologs may be excellent therapeutic targets because the modulation of these TFs affects many aspects of fat biology. We, and others, have shown that proteins function in the context of intricate molecular networks. Here, we propose to identify and study the physical and functional interactions between fat genes and the TFs that regulate their expression using high-throughput, functional genomic approaches in the experimentally tractable model organism C. elegans. By modeling these interactions into transcription regulatory networks we may be able to identify one, or a few, TFs that appear as "network hubs" and that may, therefore, play a central role in fat gene expression. Our long-term goal is to understand fat gene expression at a systems level by comprehensively mapping and studying worm fat gene networks. Here, we will test the feasibility of this using a set of 32 "core fat genes", before embarking on a larger scale project. We have chosen this set of genes because they appear to play a central role in fat metabolism. In my lab, we have all the necessary tools and methods to undertake this exciting project, including a collection of cloned worm open reading frames and a collection of cloned worm gene promoters. In addition, we recently developed a version of the yeast one-hybrid system that enables us to map fat TF-fat gene interactions in a high-throughput manner.

Our long-term goal is to decipher the mechanisms that control differential gene expression in metazoan systems. Differential gene expression is a major determinant in development, and misregulation of transcription can lead to numerous diseases. Although vast amounts of gene expression data are available, little is known about the mechanisms that regulate gene expression at a systems level. The expression of each gene is a balance between transcription activation and repression, governed by multiple transcription factors (TFs). Between 5-10% of metazoan genes encode TFs. Each TF regulates the expression of multiple target genes by binding both to protein partners and to target gene DMA. The multiple interactions TFs engage in, and the concerted action of multiple TFs per gene suggests that differential gene expression is the result of intricate transcription regulatory networks (TRNs) in which many TFs are functionally connected. The general aim of this proposal is to map TRNs that control intestinal development in the nematode Caenhorhabditis elegans. C. elegans intestinal development is an excellent system to understand differential gene expression because genome-wide intestinal expression data are available and because transcription regulation plays a pivotal role in this tissue. We will identify intestinal TRNs by mapping protein-DNA and protein-protein interactions involving intestinal promoters and TFs using high-throughput yeast one -and two-hybrid systems. We will generate a database for data tracking, data analysis and to make the data publicly available. To validate interactions, we will manually and computationally integrate protein interactions with available gene expression and phenotypic data. In addition, we will initiate the validation of protein interactions by experimental methods. The mapping of intestinal TRNs will reveal transcriptional mechanisms that underlie differential gene expression in intestinal development. Since this is a conserved biological program, these TRNs may provide insights into mammalian differential gene expression as well.

DESCRIPTION (provided by applicant): Project summary The differential expression of each of our ~25,000 genes in different tissues or under different conditions is critical for our proper development and function. Indeed, changes in differential gene expression caused by mutations in transcription factors (TFs) or the cis-regulatory genomic DNA elements they bind to (TF binding sites, or TFBSs) can result in a variety of human diseases, including several congenital disorders and cancer. In order to fully understand both normal development and pathologies, and to design effective therapeutics it is therefore critical to understand which TF regulates the expression of which gene, where and under which (developmental) conditions. In addition, it is essential to know the elements each TF binds to and where in the genome these TFBSs are located. This is a major challenge in genomic science as very little is known about the targets, binding sites, transcriptional activity and biological function for the majority of metazoan TFs. We use the nematode C. elegans as a model to address this challenge. Our long-term goal is to comprehensively map and characterize the protein-DNA interactions between all C. elegans TFs and all gene regulatory regions, and to identify all responsible TFBSs. Currently, ChIP is the most popular method to identify TF-DNA interactions. Although powerful, metazoan ChIP is limited to the few TFs that are widely and highly expressed, and for which suitable antibodies are available. To enable the identification of a wide variety of metazoan protein-DNA interactions in a condition-independent manner, we developed a high-throughput version of the yeast one-hybrid (Y1H) system. Our Y1H system can be used with several Gateway resources we created, including a promoterome that consists of 6,000 promoters, as well as ORF clones for ~80% of all 940 predicted worm TFs. Here, we propose to identify TFBSs throughout 30% of all C. elegans promoters by first mapping protein-DNA interactions between all available gene promoters and TFs by Y1H assays, and then to use these interactions to computationally delineate TFBSs. Project narrative The expression of each of our genes in different tissues or under different conditions is critical for our proper development and function. Indeed mutations in transcription factors or the genomic DNA sequences they bind to can result in a variety of human diseases, including several congenital disorders and cancer. We will identify which TF regulates the expression of which gene to gain insight into both normal development and disease, and to design effective therapeutics. Since such studies are not feasible at the genome scale in humans or mice, we have chosen the worm (C. elegans) as a model system.

DESCRIPTION (provided by applicant): What we eat greatly influences how we function and affects our propensity to get diseases such as obesity and diabetes. A major goal of Biomedical Research is to uncover which nutrients and metabolites affect our physiology on the one hand and, which of our genes mediate the physiological response to these molecules on the other. We have recently developed an innovative interspecies systems biology model of the nematode C. elegans and its bacterial diet to address these questions. We have used this model to uncover bacterially derived micronutrients and metabolites that affect gene expression, development and fertility in the worm, and to identify a C. elegans metabolic regulatory network that mediates the response to bacterial nutrients. The effects of different bacteria (and their molecules) are mediated via the animal's intestine, which functions as both a digestive and endocrine system. In the next project we will more precisely define the genes that are affected by different bacterial diets, as well as by vitamin B12 and propionic acid - molecules that are central mediators of the effects we observed. In addition, we will use systems-level phenotypic assays and genetic interaction screens to link diet-induced gene expression changes to physiological outputs such as altered developmental rate, fertility, and lifespan and the ability t response to propionic acid toxicity. Finally, we will dissect the precise mechanisms by which metabolic networks communicate with gene regulatory networks and vice versa using a set of transcription factors that are involved in mediating the response to nutritional cues.

DESCRIPTION (provided by applicant): The regulation of gene expression is vital for healthy development and physiology, and many diseases are caused by or associated with changes in gene expression. In the past decade, a tremendous amount of information has been gathered regarding transcriptional control by transcription factors that bind directly to regulatory DNA sequences in or around their target genes. In addition, microRNAs that control gene expression post-transcriptionally have been studied extensively. However, functional and biochemical information about the vast majority of RNA binding proteins has been lacking despite clear evidence of their importance in development and disease. A major factor contributing to our limited understanding of post- transcriptional control by RNA binding proteins is a shortage of appropriate technologies that start with an important mRNA, for instance corresponding to a disease gene of interest, and identify the RNA binding proteins with which this mRNA interacts. In the proposed project, we will develop a novel, high-throughput method for the detection and identification of RNA-protein interactions. We have provisionally named this technology "RNA-associated protein interaction detection" (RAPID). RAPID is based on translation and mimics endogenous RNA binding protein activity. We will first develop and apply RAPID to RNA-protein interactions in the nematode Caenorhabditis elegans because it provides a highly suitable model for further in vivo studies, and because we have clone resources such as the ORFeome available, which contains numerous full-length RNA binding protein-encoding clones. Successful completion of this project will provide the research community with a novel and broadly applicable method to detect RNA-protein interactions in an unbiased and high-throughput manner. We envision applying RAPID to the genome-scale detection of such interactions to further our understanding of complex gene regulatory networks. The methodology and RNA binding protein resource that we will develop for C. elegans will provide an important blueprint for the creation of similar resources in other model organisms and humans. PUBLIC HEALTH RELEVANCE: The regulation of gene expression is vital for healthy development and homeostasis and many diseases are caused by or associated with changes in gene expression. In recent years, tremendous progress has been made in the study of individual RNA binding proteins and how they affect gene expression. However, the genome encodes hundreds of such proteins, and methods that enable the characterization of many at one time are lacking. The proposed project is to develop a novel technology that will be broadly applicable for the functional and large-scale characterization of RNA binding proteins, and will impact both the fields of systems biology of gene expression and researchers that study one or a few disease-relevant human genes.

DESCRIPTION (provided by applicant): The transcriptional regulation of gene expression is pivotal to all biological processes. Each of our ~20,000 protein-coding genes must be expressed at the right place, time and level, as well as under the right developmental or physiological circumstances. Consequently, inappropriate gene expression is implicated in a myriad of human diseases, including congenital disorders, cancer and obesity. Transcription has been studied intensively for decades, resulting in a detailed picture of the basic biochemical mechanisms of mRNA production. However, we know little about gene regulation at a 'systems level', i.e. how TFs function together in complex gene regulatory networks (GRNs) to faithfully orchestrate the expression of large sets of genes. Our long-term goal is to comprehensively characterize the structure, function and evolution of complex metazoan GRNs to gain insights into global mechanisms of gene regulation. It is becoming increasingly clear that textbook explanations of gene regulation in which a TF binds DNA in the genome and upon doing so regulates the most proximal gene are too simplistic because many physical TF binding events lack an apparent regulatory consequence. There are several explanations for this, ranging from technical (e.g. detection limits, attribution of a bindng event to the wrong gene) to biological (e.g. redundancy between TFs, condition-dependent effects). Conversely, regulatory interactions are not necessarily due to a direct effect. For instance, TFs can function in cascades to propagate functional regulation. A major challenge is to combine physical and regulatory interactions to increase our understanding of the mechanisms of gene regulation in the context of complex multicellular organisms. Many GRN studies focus either solely on physical TF interactions, whereas others focus primarily on regulatory interactions. However, integrated GRNs that combine both are scarce and, when available are relatively small in scale. If we had high-quality, large- scale physical and regulatoy interaction data, as well as spatiotemporal and conditional gene expression data, we could build increasingly precise GRNs. Here, we will continue our studies on the nematode C. elegans to map and integrate physical and regulatory GRNs, which will help us to go beyond mapping to understanding the regulatory mechanisms of gene expression at a systems level in a complex multicellular organism.

? DESCRIPTION (provided by applicant): As we use the vast genomic and epigenomic knowledge obtained over the past fifteen years to understand health and disease, there is a growing realization that we need to develop a quantitative understanding of the components of various cell types and tissues in terms of their concentrations and interactions. This knowledge is necessary to build predictive dynamical models to connect the genomic and epigenomic characteristics to biochemical and physiological functions. To obtain this knowledge it will be necessary to experimentally identify and determine concentrations all proteins (and other cellular components) along with the kinetic parameters that govern their interactions in all cell types from major organs of the human body. This quantitative and dynamic experimental cataloging of proteins and their interactions will be anchored in the human genome and transcriptome at one end, and human cell physiological responses at the other. The quantitative data will be most useful when organized in a database that enable building of dynamical ODE and PDE models as well as network models. To ensure reproducibility and wide usage of these kinetic parameters the metadata describing experiments should be both deep and broad. To develop a bottom-up community driven plan for a Human Quantitative Dynamics (HQD) Project we are seeking partial support for a workshop for which bring together researchers from different communities who would generally not meet with one another. This would include molecular neurobiologists, cellular immunologists, gastrointestinal biologists, chemical and mechanical engineers, bioinformatics researchers and dynamical modelers. The workshop will be held at the NIH in Bethesda and will be open and free. At this workshop these researchers will analyze the strengths, weaknesses and cost /benefit ratios of recently completed large scale projects; evaluate the current level of knowledge, identify the contours of a plan for a pilot project and for future workshops that can develop and assess more in-depth plans of how we can come together to develop resources and technologies that can enable quantitative and predictive understanding of human cell biology and physiology.

DESCRIPTION (provided by applicant): Life history traits are closely associated with metabolic processes. For example, in the model organism C. elegans, changing the diet from one bacterial species to another or perturbations in certain metabolic genes can significantly change fecundity, life span, and the rate of development. Therefore, C. elegans is an excellent animal model for understanding the relationship between metabolically mediated phenotypes and dietary or genetic manipulations. Gaining such understanding requires a systems-level reconstruction of its metabolic network as well as an experimental platform to monitor its conversion of nutrients to biomass and energy. Global metabolic network models are available for more than 50 organisms, mostly prokaryotes but also eukaryotes such as Saccharomyces cerevisiae, Arabidopsis thaliana and Homo sapiens. In prokaryotes and yeast, reconstructions have been combined with experimental measurements and predictions of growth rates and other phenotypes in bioreactors. Most intact multicellular organisms cannot be grown under precisely controlled conditions that mimic a prokaryotic bioreactor. C. elegans, however, has several distinctive properties that uniquely enable this, including short life span, hermaphroditic reproduction, and simple morphology. In this project, we will first reconstruct a genome-scale metabolic network model for C. elegans and subsequently calibrate and validate this model experimentally. We will initiate the compartmentalization of the model into three parts: the intestine (the major metabolic organ), the germline (the reproductive organ that generates biomass), and the other somatic tissues. We will calibrate the model with experimental parameters, including biomass composition, food uptake rates, and maintenance ATP to define growth in liquid cultures. Finally, we will validate our model with independent tests, including dietary and genetic manipulations with expected phenotypes that are to be predicted by the model via biomass production and other metabolic rates. The worm model proposed in this study, together with the quantitative liquid culturing techniques, will serve as a unique toolbox for the analysis and genetic engineering of C. elegans and create opportunities for a broad array of applications at a systems level.

Biological networks describe relationships between biomolecules. We study two types of biological networks and the interactions between them. First, we characterize physical and functional gene regulatory networks that are primary controllers of gene transcription during development and growth, to maintain homeostasis and to respond to environmental cues and insults. We will expand our studies on physical interactions between gene promoters and transcription factors in C. elegans to focus on true functional, tissue-specific regulatory networks in vivo. This will involve the examination of non-transcription factor regulators that affect transcription indirectly. Second, we study metabolic networks that convert nutrients into biomass and energy. We will perform experiments to gain insight into how metabolic networks are wired in different tissues and under different nutritional conditions. In addition, we will continue to develop our WormFlux website to enable data integration and flux balance analysis with selected parameters such as biomass composition and different objective functions. Finally, and importantly, we study how gene regulatory networks affect metabolic networks and vice versa. We will test how intestinal transcription is affected by perturbations in metabolic genes. In addition, we will generate gene expression data under different conditions to examine how gene expression changes influence metabolism. Our data will provide broad phenomenological and deep mechanistic insights, as well as a set of resources for the larger community.

SUMMARY Individuals can respond to diverse nutrients and dietary restrictions in markedly different ways. Some people easily gain weight, but others remain thin no matter what they eat. Additionally, metabolic diseases can differ dramatically among individuals in a population, for both rare single-gene Mendelian diseases and common multifactorial metabolic diseases such as obesity and type 2 diabetes. In large part, this variability suggests that individual genetic differences greatly affect the likelihood to get sick as well as the severity of the illness for both rare and common metabolic diseases across a population. It would be extremely valuable if one could identify both rare and common variants that contribute to individual responses to diet and to the acquisition of different types of metabolic diseases. Rare variants are usually identified by linkage mapping and whole- genome sequencing using families with affected individuals. By contrast, common variants are usually identified by genome-wide association studies using large populations of people with and without a disease. We will develop personalized metabolic network models for a large set of genetic individuals of the nematode C. elegans, both representing healthy metabolic state and mimicking an inborn error of human metabolism. With our experimental system and approach we will be able to derive predictions of both rare and common variation in a variety of metabolic traits influenced by nutrition. We will extensively validate such predictions using CRISPR/Cas9-mediated genome editing.