Project Summary/Abstract In development, the embryo generates many cell types with distinct gene expression programs, leading to heterogeneity across cells. In cancer, mutation generates heterogeneity, with a growing recognition that non- genetic (epigenetic) mechanisms contribute to tumor heterogeneity and treatment failure. The ability to take the same genetic template and form different cell states, exhibiting cell-to-cell variation in gene expression and behavior, is fundamental to many cell systems across problems in human health including stem cells, cancer, and immune cell function. Yet, the mechanisms by which cell systems encode heterogeneity in gene expression are unclear. Pluripotent embryonic stem cells give rise to all the cells of the adult mammal, and primary embryonic stem cells are genetically stable in culture, making them an ideal system for studying the emergence of non-genetic heterogeneity in cell systems. Cell-to-cell variation in gene expression that arises in the absence of environmental signals has previously been termed `noise', or attributed to stochastic processes of gene expression. However, gene expression heterogeneity is reproducible, suggesting regulation. The goal of this proposal is to identify regulatory mechanisms by which the genome encodes non-genetic heterogeneity in cell systems. We will use embryonic stem cells as a model system for gene expression heterogeneity. First, we will identify transcription factor pairs whose combined activity at enhancers leads to transcriptional heterogeneity of the regulated gene. In order for heterogeneity to result in forming distinct states with the potential to prime stem cells into different lineages, heterogeneity must be heritable. Second, using a modified Luria-Delbruck fluctuation analysis, we will identify memory loci capable of heritable transmission of non-genetic heterogeneity across cell generations. Motifs such as autoregulation, whereby a gene's product can regulate its own production, may contribute to the formation of different cell states and therefore to heterogeneity across a cell population. Third, we will manipulate a single factor within or outside of an autoregulatory loop to determine how it impacts heterogeneity. To accomplish these goals, we will leverage an assay we have developed which allows the identification of actively transcribed regulatory regions and genes in small subsets of cells. Completion of these goals will address longstanding questions about the origins of heterogeneity in cell systems and will advance a systems level approach for understanding cell state. If successful, the proposal may unlock future studies applying similar approaches to identify non-genetic drivers of tumor heterogeneity and treatment failure. Therapeutic targeting of these mechanisms may unlock new treatment strategies for cancer.