Environmental regulation of estrogen responsive genes in single living cells

National Institute Of Environmental Health Sciences

Investigators

Linked publications, trials & patents

Abstract

Our group is interested in somatic gene regulation in disease. Over the past decade several groups have characterized significant expression heterogeneity in stem cells, progenitors, and terminally differentiated cells across several tissues. Although the function of this heterogeneity in normal tissue has not been extensively explored, heterogeneity enables cancer to evade chemotherapeutic treatments. Moreover, genes involved in resistance and metastasis are transiently expressed. This dynamic gene expression highlights the importance of identifying dysregulated mechanisms of non-genetic heterogeneity to improve treatments. A great example is heterogeneity in estrogen receptor-alpha positive (ERa+) breast cancer which makes up over 70% of diagnosed breast cancers. Like other cancers, tumors are very heterogeneous in ER+ expression. Some tumors contain as low as 1% ERa+ expressing cells. Since the estrogen receptor alpha (ERa) is a transcription factor which drives several functions including cellular proliferation, targeting its response has been the primary mode of treatment. These cancers are effectively treated with ERa antagonists such as tamoxifen. However, resistance to prolonged tamoxifen treatment is common and often involves mutations in ERa, altered signaling pathways and epigenetic reprogramming. Intriguingly, knockdown of a chromatin remodeler reduced expression heterogeneity of ERa+ breast cancer and led to more sensitivity to ERa antagonists. Most of our molecular understanding of the estrogen response comes from cancer cell line studies. Within 40 minutes after the addition of estradiol, hundreds of genes are activated and repressed. This partly achieved through estrogen receptor and cofactor recruitment to distal regulatory sequences called enhancers. These sequences function through local proximity, although how exactly they aid in transcriptional initiation is unclear. Although we have a genome wide view of the estrogen transcriptional response, single cell studies have shown that only a fraction of hormone responsive genes within single cells respond to hormone. These data suggest that there is either a rate limiting factor needed for transcription initiation of hormone responsive genes or that individual gene loci exist in stable transcriptionally permissive and non-permissive states. Forty years ago, electron micrographs of Miller chromatin spreads illustrated that genes toggle between active periods of nascent RNA synthesis and periods devoid of RNA. These transcription units or transcriptional bursts represent the culmination of multiple regulatory processes. Therefore, analyzing how often, regularly and long these bursts occur can provide regulatory insight. To understand the biochemistry of bursting is to understand how transcription works in the nucleus. Several groups have imaged transcriptional bursts in live cells using fluorescently labeled MS2 and PP7 proteins which bind strongly to RNA stem loops. These RNA stem loops are inserted into transgenes or at endogenous loci, followed by integration of stably expressed MS2-GFP protein. As soon as the gene and MS2 RNA stem loops are transcribed, several MS2-GFP molecules the nascent RNA, and a bright punctate spot denoting an active transcription site is observed. As a postdoc, I used this system to investigate the variably expressed TFF1 gene. I observed that it exhibited variably long periods of inactivity: lasting minutes to days. These periods of inactivity completely explained TFF1 expression heterogeneity. In a separate study we extended this observation to several genes involved in other pathways. Combined these data indicated that genes transition through dynamic states from deep repression to high transcriptional activity. However, work from lineage tracing experiments involving monoclonally expanded T cells indicate that for a small number of genes, the decision to express is permanently set and maintained after differentiation. In other words, expression of these genes is stably active or inactive throughout hundreds of cell divisions. In this case, heterogeneity is stable and is not a dynamic process in terminally differentiated cells. This discrepancy between our results and these studies has several possibly explanations including specialized genes and cell types. However, in normal tissue its difficult to consolidate stable expression heterogeneity, which is stochastically set by a dynamically expressing progenitor, and epithelial cell replacement. The simplest explanation is that these models are all part of the same gene paradigm with a continuous spectrum of transition rates. Specialized genes or cells have a different balance of how much time genes are active and inactive. As we age, epigenetic drift is enhanced through exposure and mutations in chromatin remodelers and transcription factors. These perturbations shift the balance both globally and at susceptible gene loci. Therefore, our approach must look at both normal and disease tissues to understand heterogeneity. Our goal in this project is to characterize the diversity in enhancer and promoter gene states. This project will help us determine regulatory features of highly bursting or deep repressed genes by grouping genes with similar transcriptional bursting characteristics. We have characterized enhancer and promoter activity and observe that enhancers are more sensitive to estradiol than their cognate target promoter. These data suggest that promoters exhibit thresholds which enhancers must help overcome before the gene is transcriptionally active. Additionally, we have made significant progress developing reagents to visualize transcription of enhancer and promoters in live cells and in whole mouse mammary glands. These reagents will help us visualize enhancer and promoter states. Lastly, we have characterized how low dose endocrine disruptor chemicals alter estrogen sensing directly, and through enhancer reprogramming. Our work has implications into how chronic low dose endocrine disruptor chemical exposure alters transcriptional responses and heterogeneity of tumors samples. In addition to this work, we published work with our collaborators describing how elements of super-enhancers can have either activating and inhibitory roles.