Hematopoietic Stem Cell Biology

National Human Genome Research Institute

Investigators

Linked publications, trials & patents

Abstract

There are two related aims in this Project. Aim 1. Single cell transcriptional profiling of primitive mouse hematopoietic cell populations. We have determined that transcriptional and epigenetic profiles are determined for megakaryocytic, lymphocytes, neutrophils and monocytes primitive populations of hematopoietic progenitor cells before they have committed to their specific lineages. We have developed methods to prospectively isolate these biased subpopulations for more in-depth study. Using the prospectively isolated subpopulations in we will define their epigenetic profiles. We will determine the chromatin accessibility as well as the DNA methylation and modifications (methylation, acetylation) to histone proteins across the genome to correlate with mRNA expression and colony formation in vitro. Aim 2. Comparison of the transcriptional and epigenetic profiles of differentiating mouse hematopoietic cells to the available data in humans gathered as part of ENCODE. In mammals, the number of HSC per animal is remarkably similar between species despite great differences in the number of cells produced. Hematopoiesis has traditionally been modeled as a hierarchy in which the progeny of HSC become progressively committed to a branch (myeloid or lymphoid) and ultimately to a single lineage in a stepwise fashion. Recent single cell analyses and improved colony assays have shown that hematopoiesis follows a more fluid process where lineages are specified (at least at the transcriptional level) earlier in hematopoiesis than the original hierarchical model predicted. For example, we found that the human MEP population contains three subpopulations of lineage primed cells that could be prospectively separated by surface marker expression. In Aim 1 we have completed a single cell profile of over 10,000 cells each from primitive hematopoietic stem and progenitor cells (LSK), Common Myeloid Progenitors (CMP), Megakaryocyte/Erythroid progenitors (MEP) and Granulocyte/Monocyte progenitors (GMP). In these populations we have identified 25 distinct transcriptional profiles. We performed indexed single cell analyses on the CMP population to link these profiles to cell surface markers for prospective isolation by flow cytometry. We identified and isolated 4 subpopulations of CMP cells, each with a distinct transcriptional profile: Mixed progenitor, megakaryocyte, megakaryocyte/erythroid and granulocytic. The prospectively sorted cell populations are multipotent in methylcellulose assays despite their transcriptional similarity to specific lineages. In liquid culture, the colony output of each population is skewed towards the transcriptional profile. In a collagen based assay that favors megakaryocytic colony formation, the granulocytic population is not capable of megakaryocytic colonies. Using the Assay for Accessible Chromatin (ATAC) on the 4 bulk subpopulations of CMP, we have shown that each cell population has a unique chromatin architecture. We are now performing the ATAC on single CMP cells to determine whether the chromatin accessibility is as uniform as the transcriptional profiles. We are correlating the changes in the chromatin with the gene expression profiles and colony formation to determine the first steps of hematopoietic differentiation. Aim 2. The comparison of the transcriptional and epigenetic profiles of differentiating mouse and human hematopoietic cells is well beyond the capacities of any single lab. Thus we have joined forces with both ENCODE to provide subject expertise and sorted human cells for analysis as well as the ValIdated Systematic IntegratiON of hematopoietic epigenomes consortium (VISION). VISION has focused on the mouse system for several reasons. Mouse and human hematopoiesis, while highly conserved in some respects, also differ in many significant ways. By comparing the mouse and human epigenetic profiles, we will identify overlapping (common) patterns as well as distinct patterns that can be associated with the different properties of mouse and human erythropoiesis, which will generate more informed hypotheses than would be possible by studying hematopoiesis in a single species. For example, haploinsufficiency of ribosomal proteins in humans leads to a block in erythropoiesis resulting in Diamond Blackfan anemia, while haploinsufficiency of ribosomal proteins in mice is benign. We believe that identifying the differences (for example) in mouse and human RP gene regulation may identify a mouse like pathway that could be targeted to treat DBA patients. Our group has developed transcriptional, transcription factor binding, DNA methylation and histone modification profiles that complement data generated in the laboratories of Mitchell Weiss (St. Jude Childrens Research Hospital), Gerd Blobel (University of Pennsylvania), Jim Hughes (Oxford University) and Doug Higgs (Oxford University) to define the chromatin landscape during hematopoietic differentiation. These data have been integrated into a comprehensive 3D profile by the computational arm of VISION, which includes Jens Lichtenberg from our group, Berthold Gottgens (University of Cambridge), James Taylor (Johns Hopkins), Feng Yue (Penn State Hershey), Yu Zhang (Penn State) and Ross Hardison (Penn State). We have shown that the epigenetic profile of megakaryocytic is established in the most primitive hematopoietic stem and progenitor cells and is maintained during differentiation. In contrast, the epigenetic profile of erythroblasts is acquired during differentiation. We have found that the chromatin state profile varies from lineage to lineage and locus to locus. We are testing the hypothesis that each lineage has a specific epigenetic master regulator that is responsible for instituting the lineage specific program.