Robi D. Mitra - US grants

[unreadable] DESCRIPTION (provided by applicant): The proposed experiments aim to fill a major gap in our ability to map protein-DNA interactions: we can identify the target genes of any DNA-binding protein (by the "chromatin immunoprecipitation" method), but we cannot do the converse: identify all the proteins that bind to a given promoter. We will attempt to gain this capability for the yeast S. cerevisiae by developing an in vivo method for mapping DNA-protein interactions. The method employs the Ty5 retrotransposon, which will be engineered to provide each DNA-binding protein of yeast with a "calling card" that it deposits at regions of the genome where it binds. Recovery of these "calling cards" and determination of the information they contain (a DNA sequence identifier of the protein they are associated with) will reveal the proteins that visit particular regions of the genome. We will attempt to employ a high-throughput DNA sequencing method that promises to enable mapping of the sites of binding of all yeast transcription factors in one experiment. We intend to apply the method to produce a transcription factor-gene interaction map for yeast grown under a variety of conditions, which should contribute to defining all functional elements of the genome of this experimental organism that provides a model for human biology and disease. Proper regulation of gene expression is necessary for normal development of organisms; abnormal gene expression can lead to diseases such as cancer. Regulation of gene expression is effected by DNA-binding proteins that bind near genes, and to understand this process and why it goes wrong we need to know which DNA-binding proteins bind near which genes. We propose to develop a facile method for doing this in the yeast S. cerevisiae, a popular organism for experimentation that provides a model for human biology and disease, and if successful, may be able to be adapted for experimentation with other organisms, including humans. [unreadable] [unreadable] [unreadable]

DESCRIPTION (provided by applicant): Epigenetic marks such as DNA methylation play important roles in mammalian development and the onset of disease. Our ability to study methylation in complex tissues is limited because existing tools measure the average methylation of a large number of cells. Since most tissues contain many different cell types, these bulk measurements are inaccurate. We hope to create an entirely new way to analyze methylation in complex tissues. We propose MethylMap, a high-throughput technology that combines multiplex amplification, sample-specific DNA barcodes, and next- generation sequencing to analyze methylation in laser capture micro-dissected cells. Our research plan consists of two specific aims. First, we will demonstrate the multiplexed amplification and single-molecule sequencing of 5000 promoters from a single sample of bisulfite treated DNA. We will use this technology to catalogue epigenetic mutations that arise during tumorigenesis. Next, we will use the 454 FLX sequencing platform to bisulfite sequence 200 promoters from laser capture microdissected cells arrayed in 96 well plates (or 19,200 promoters per plate). Each sample will be barcoded with a DNA tag to connect the measured methylation pattern with the location of the cells. We will build on preliminary results that demonstrate 1) multiplexed amplification of 90 loci from a single sample, 2) deep single molecule bisulfite sequencing of the MLH1 promoter in tumors using the 454 Life Sciences FLX sequencer, and 3) the use of sample-specific DNA barcodes with next-generation sequencing technology. MethylMap will have many applications: it will be used to molecularly classify cells in complex tissues, to determine the lineages of somatic cells, and to probe the role of aberrant methylation in tumorigenesis. We believe MethylMap will become an important tool for scientists interested in development, tissue homeostasis, and tumor biology. PUBLIC HEALTH RELEVANCE: DNA methylation is an important mechanism used by the cell to control gene regulation. Recent studies have shown that DNA methylation is critical for mammalian development, and if DNA methylation is misregulated, it can lead to cancer. Current tools to study DNA methylation are inadequate because they cannot effectively analyze complex tissues that contain many different types of cells. We propose to develop MethylMap, a technology that will make it possible to analyze the methylation patterns of cells in complex tissues for the first time. This tool will provide new insights into how tissues form, and how cancers arise.

DESCRIPTION (provided by applicant): Most biological processes are executed by proteins, but no method currently exists to accurately measure protein abundance and post-translational state proteome-wide. To redress this deficiency, we propose Digital Analysis of Proteins by End Sequencing (DAPES), a method that sequences many individual peptide molecules in parallel using Edman degradation. DAPES will be cost-effective, highly sensitive, and quantitative. DAPES is based on two innovations - 1) the use of dye-labeled antibodies to inexpensively and robustly detect single peptide molecules; and 2) a strategy that uses a universal set of ~20 antibodies to sequence peptide molecules. Our previous work, in which we used fluorescent antibodies to detect and quantify protein levels by single molecule counting, demonstrates that this approach is realistic and powerful. PUBLIC HEALTH RELEVANCE: Most cellular functions are performed by proteins, yet current methods are unable to accurately quantify protein levels and post-translational state in a comprehensive manner. This shortcoming is preventing a quantitative understanding of normal cellular processes, the mechanisms by which they fail, and how these failures lead to disease. To redress this deficiency, we propose to apply recent advances in single-molecule imaging to the field of protein detection. By sequencing single peptide molecules in parallel we will develop a protein analysis tool with unprecedented sensitivity, dynamic range, and utility.

DESCRIPTION (provided by applicant): Transcriptional networks that control the development of organisms are precise, highly coordinated, and complex. This proposal seeks to understand how the same transcription factor can specify multiple distinct cell fates during the development of an organism. Specifically, we are interested in the transcription factor Olig2, which can promote both a motoneuron and an oligodendrocyte cell fate. We hypothesize that Olig2 is able to perform its multiple functions through interactions with other DNA binding proteins causing it to bind different targets in different cellular contexts. Testing this hypothesis using current methods for the analysis of transcription factors is difficult because they cannot trace binding throughout a cell lineage, making it impossible to correlate DNA-binding events in progenitor cells to the final cell fates of their progeny. We propose to use a novel method, transposon Calling Cards, to record Olig2 binding during neural differentiation. The method entails fusing the transposase of a transposon to a transcription factor, thereby causing it to direct the insertion of transposon DNA into the genome near where it binds. The transposon becomes a Calling Card that permanently marks the transcription factor's visit to that place in the genome. By recovering these Calling Cards along with some of the genomic DNA that flanks them and then determining their DNA sequences, it is possible to map the genome-wide binding history of the transcription factor. We propose to apply the Calling Card method to understand how Olig2 carries out its distinct functions. Since many important transcription factors perform more than one function during development, the insights that we gain from this work should be broadly applicable. Our specific aims are 1) to trace Olig2 binding through neural stem cell differentiation to understand how it promotes two distinct cell fates, 2) to confirm that differentially bound target genes promote motoneuron or oligodendrocyte cell fates, 3) to analyze Olig2 and Ngn2 binding in living zebrafish using Calling Card technology. These aims are feasible: we have successfully implemented the Calling Card method in both yeast and mammalian cells, and our preliminary results demonstrate transcription factor directed insertion of Calling Cards in zebrafish. We are confident that the rewards of further developing this technology and applying it to understand the process of cell fate specification will be substantial

DESCRIPTION (provided by applicant): Schwann cell transplantation holds great promise for the treatment of spinal cord injuries and some neuropathies. In addition, Schwann cell functions are coming under wider scrutiny due to their potential importance in hematopoiesis. A major bottleneck hindering the progress of Schwann cell-based therapy and Schwann cell functional genomics is the lack of methods to produce large numbers of transplantable cells and the easy perturbation of their genetic network. Recently, it has become possible to reprogram fibroblasts into different cell types by expressing a small number of transcription factors. However, the efficiencies are typically low, and only a few cell types (e.g. neurons, cardiomyocytes, oligodendrocytes) have been produced to date. We propose to overcome these difficulties by creating artificial transcription factors (ATFs) based on the Cas9 protein. Cas9 can be directed to bind specific genomic sequences using guide RNAs, so it will possible to specifically activate hundreds or even thousands of genes. We will use Cas9 ATFs to reprogram fibroblasts into neurons and Schwann cells by activating transcription factors that are specific to these cell types. We anticipate that this approach will substantially improve the efficiencies of existing transdifferentiation protocols (for conversion into neurons), as well as enable transdifferentiatio to previously unobtainable cell types (Schwann cells). Our preliminary experiments suggest our strategy is feasible. We have demonstrated that Cas9 ATFs can achieve potent gene activation (>100 fold), and we have developed computational methods to predict the sets of genes required for transdifferentiation. Our specific aims are as follows: 1) To determine the rules that govern gene activation by Cas9-based artificial transcription factors (ATFs). 2) To develop tunable Cas9 mutant proteins bearing transcriptional activation or repression domains wherein their activity can be controlled by addition of small molecules to enable regulable perturbation of large-scale genetic networks. 3) To transdifferentiate fibroblasts into Schwann cells or their precursors by simultaneously activating the expression of 75-100 transcription factors that are differentially expressed between these two cell types.

? DESCRIPTION (provided by applicant): The brain is a remarkably complex organ comprised of hundreds of unique cell types that are organized to form sophisticated neural circuits. Although we have made progress toward understanding brain function and development, it is clear there is still much to be learned. Currently, all genome-wide methods that could be brought to bear on functional studies of the brain are destructive, meaning that as a genomic analysis is performed on a population of cells, the cells are destroyed. This fact limits our ability to connec early molecular events in the cells of the brain with later behavioral or cellular changes. For example, it is currently impossible to connect transcriptional changes in a neuron with knowledge of whether or not the cell was successfully incorporated into a memory trace. Similarly, it is not feasible to connect the early molecular events that occur in a neuronal progenitor cell with the final cell fate decision made by the cell. We have set out to develop a transformative technology that can record molecular events at the time that they occur and can then be read out later after any defined period of time. We have a novel technology called transposon `Calling Cards' that, in culture, provides cells with a molecular memory of protein-DNA interactions that occur at a particular moment in time. Here, we propose to adapt this technology for use in vivo enabling a retrospective genomic analysis of molecular events. We will demonstrate the utility of this technology by completing four test-case experiments that cannot be done with existing methods. Specifically, we will test the method by: 1) retrospectively identifying candidate transcription factors that control the specification of cell types in the CNS 2) identifying features that distinguish neurons resistant to neurodegeneration in vivo, 3) identifying the neurons that become active during mouse vocalization behavior while simultaneously mapping the genome-wide binding of activity-dependent transcription factors in these neurons, and 4) identifying the molecular features that distinguish neurons that were incorporated into a fear memory trace from those that were not incorporated.

Project Summary The reprogramming of skin or other easy to obtain cell types into therapeutically useful cells holds great promise for regenerative medicine, but the practical usefulness of this technology has been limited because typically only a small percentage of cells are successfully converted to the target cell type and these are often immature or incompletely specified. Single-cell expression analysis has revealed that most reprogrammed cells start with very similar expression profiles but then acquire a wide variety of different fates, many of which are developmental dead ends. Why does a cell population that is initially homogenous produce a myriad of different cell fates, with only a few cells achieving the targeted fate? This question is extremely difficult to answer with existing methods. The main focus of our proposal is to develop self-reporting Calling Cards, a new technology that can answer this question by simultaneously measuring transcription factor binding and genome-wide mRNA levels from thousands of single cells. We will demonstrate the utility of self-reporting Calling Cards by mapping the binding and function of the pioneer factor Foxa1 and its cofactor Hnf4a during the reprogramming of fibroblasts into induced endoderm progenitors (iEPs). We hypothesize that the direct targets of these TFs are stochastically expressed in cells undergoing lineage reprogramming, and that by forcing the expression of target genes that are usually transcribed only in successfully converted cells, we can improve overall target cell yield and maturity.

Transcription factors shape gene expression by binding to genomic cis-regulatory elements and then recruiting nucleosome remodeling factors, the RNA polymerase holoenyzme, and other transcriptional coactivators. What determines where a transcription factor binds in vivo? For most eukaryotic transcription factors (TFs), the answer to this question is not known. We have recently analyzed two yeast bHLH proteins, Cbf1 and Tye7, which have nearly identical DNA binding preferences in vitro, but bind at almost completely non-overlapping target loci in vivo. We found that Cbf1 utilizes homotypic cooperativity to achieve its specificity, while Tye7 binds in a TF collective, a phenomenon that has been described only recently in Drosophila, but is poorly understood. We hypothesize that homotypic cooperativity and collective binding are widely used by eukaryotic TFs to achieve their specificities in vivo. We will test this hypothesis by quantifying the contribution of these two mechanisms to the in vivo binding of all yeast transcription factors. We will also dissect a small number of these complexes in detail. We will investigate the binding specificity of the human bHLH transcription factor Usf1, which is a candidate drug target because of its involvement in obesity and metabolic disease. The completion of this work will deepen our understanding of the factors that govern the in vivo specificities of transcription factors. Furthermore, by gaining an understanding of the protein-protein contacts that regulate Usf1 binding, we will uncover interactions that can be disrupted for therapeutic benefit.

PROJECT SUMMARY The brain is the most complex organ in the body, consisting of hundreds of molecularly, physiologically, and anatomically distinct cells. Recently, methods have been developed that can cost-effectively measure mRNA abundance in tens of thousands of single cells, and this has led to a revolution in the identification and classification of new types of cells in the brain. But these methods only measure one aspect of gene regulation ? mRNA levels. To fully understand the transcriptional networks that function in the brain, it will be important to also map the genome-wide binding locations of transcription factors and chromatin modifiers in the myriad of different cell types present in the brain. We propose to develop a method to simultaneously map transcription factor binding and measure mRNA abundance from single cells in the brain. To do so, we will adapt our transposon based methods for measuring the binding of DNA interacting proteins to existing single cell profiling methods. This technology, single-cell Calling Cards, builds on our previously developed transposon Calling Card method, but significantly extends the method allowing use in populations of heterogeneous cells, without a priori definition of cell type. We propose here to develop new mouse lines compatible with the wide range of existing resources in mouse, as well as viral and plasmid reagents applicable across model species, and distribute these to the community. Finally, since the tools we will develop generate new types of data, we will develop a user- friendly software for data visualization and analysis of TFs or chromatin modifying proteins binding data across multiple cell types. Robust resources for analysis are crucial if these technologies are to find broad use in the community. Our primary goal is to enable the parallel analysis of transcription factor binding and mRNA expression levels from tens of thousands of single cells in the brain. !

PROJECT SUMMARY The brain is the most complex organ in the body, consisting of hundreds of molecularly, physiologically, and anatomically distinct cells. Recently, methods have been developed that can cost-effectively measure mRNA abundance in tens of thousands of single cells, and this has led to a revolution in the identification and classification of new types of cells in the brain. But these methods only measure one aspect of gene regulation ? mRNA levels. To fully understand the transcriptional networks that function in the brain, it will be important to also map the genome-wide binding locations of transcription factors and chromatin modifiers in the myriad of different cell types present in the brain. We propose to develop a method to simultaneously map transcription factor binding and measure mRNA abundance from single cells in the brain. To do so, we will adapt our transposon based methods for measuring the binding of DNA interacting proteins to existing single cell profiling methods. This technology, single-cell Calling Cards, builds on our previously developed transposon Calling Card method, but significantly extends the method allowing use in populations of heterogeneous cells, without a priori definition of cell type. We propose here to develop new mouse lines compatible with the wide range of existing resources in mouse, as well as viral and plasmid reagents applicable across model species, and distribute these to the community. Finally, since the tools we will develop generate new types of data, we will develop a user- friendly software for data visualization and analysis of TFs or chromatin modifying proteins binding data across multiple cell types. Robust resources for analysis are crucial if these technologies are to find broad use in the community. Our primary goal is to enable the parallel analysis of transcription factor binding and mRNA expression levels from tens of thousands of single cells in the brain. !

Abstract: This proposal seeks the renewal of the Opportunities in Genomics Research (OGR) Program. Since 2007, the OGR Program has been run out of The McDonnell Genome Institute at Washington University in Saint Louis and funded through the NHGRI's Diversity Action Plan. The mission of the OGR Program is to increase the representation of students from underrepresented groups in genome science or genome science-related PhD and MD/PhD programs. We seek to accomplish our mission through the effective execution and evolution of summer undergraduate and academic year post-baccalaureate research programs for students from racial and ethnic backgrounds underrepresented in the sciences, students from disadvantaged backgrounds, and students with disabilities. Both OGR programs provide trainees with cutting-edge research experiences in genome science or related fields and 11 other educational activities that seek to train them to think critically and to write and speak effectively about their research. Workshops and classes are tailored to train students in the core concepts of genome science, bioinformatics, and scientific presentation. Professional development activities focus on helping students prepare for and excel in graduate school interviews and graduate school itself. Over its three funding cycles, student outcomes for the OGR program reveal great momentum. PhD matriculation of students in our summer program rose from 29% in the first cycle to 50% or above the last two cycles. More impressively, PhD matriculation of students in our post-baccalaureate program rose from 44% in the first cycle to 75% in the second cycle and 90% in the current cycle. We expect the implementation of new and modified activities to solidify and extend these gains. Specifically, new activities will promote integration among OGR trainees and trainees in the NHGRI T32 PhD Genomic Sciences program, enhance training in bioinformatics, provide trainees with greater exposure to the diversity of research careers available to PhD scientists, and initiate community outreach events to under-served area high schools and their students. Based on the past success and current plans of our program, we seek support for five more years of funding to continue our mission of enhancing diversity in PhD programs nation-wide. We propose to expand our post- baccalaureate program to five trainees per year, and with university support, to maintain the size of our summer undergraduate program at eight students per summer. We believe our programs will continue to help realize the great scientific and intellectual potential inherent within the diverse population of the United States, much of which currently lies latent due to the underrepresentation of many sectors of our nation's population within scientific research in general and genome science in particular.

Abstract The brain is remarkably complex, and our understanding of this organ is still in its infancy. Many fundamental questions about brain development and function remain. Single cell genomics promises new ways to answer these questions, but nearly all single-cell methods share shortcoming ? cells are destroyed when their molecular states are measured. This ?destruction upon observation? makes it impossible to correlate molecular events that have occurred in the past with the final outcomes: if you destroy an intermediate progenitor in the brain to profile its epigenome, you have no way to know what cell fate the progenitor would have adopted. Likewise, profiling an animal after an experience is going to conflate the molecular consequences of the experience with any molecular predispositions that were there before exposure. And if you profile before the experience, you have no way of know of the animal's outcome ? e.g. if the animal would have been a learner or not. This shortcoming is inherent to nearly all existing genomic methods and impedes a wide variety of interesting analyses. We have recently developed a platform technology, single-cell `Calling Cards' (CC) that uses transposons to capture molecular events at one instant in time and then read them out at a later time with single cell resolution. We now propose to build upon this foundation to develop a robust, easy to use technology platform to record molecular events at single cell resolution in the mouse brain and connect them with cell fate decision, behavioral outcomes, neuronal activity profiles, and anatomical location. We will develop a set of widely applicable reagents and methods, apply them to a series of high-impact ?test cases? to demonstrate their utility, and rapidly distribute them to the broader community.