Our lab studies the mechanisms that dictate the development of hematopoietic and cardiovascular stem/progenitor cells, with the ultimate goal to generate these cells in culture for therapeutic use. We integrate mouse models, human HSC, in vitro differentiation of ES cells and state-of-the-art molecular and functional assays to pursue this goal. We have made novel discoveries in transcriptional mechanisms and niche signals that regulate mesoderm divergence to hematopoietic and cardiac stem/progenitor cells, and protect HSCs in the placenta and fetal hematopoietic niches.
The development of blood cells and cardiovascular system is intimately linked both through their ancestry and function. Blood, heart and vasculature share a common origin in Flk1+ mesoderm, and a defect in one component can impair the development and function of the other. Although individual regulators for blood and cardiovascular lineages have been identified, we lack knowledge of the roadmap of how to generate stem/progenitor cells for each tissue, and how the lineage specific master regulators exploit and modify the epigenetic landscape to establish cell fates.
Scl/Tal1 induces hematopoietic fate and represses cardiomyogenesis in embryonic endothelium:
Recent studies verified that hematopoietic stem/progenitor cells (HS/PCs) emerge from endothelial precursors during embryogenesis; however, little was known how specific endothelia gain hemogenic competence. We showed that bHLH transcription factor confers hemogenic identity for endothelium by activating a broad HS/PC transcriptional program (Van Handel, Montel-Hagen at al. Cell 2012).. Moreover, we uncovered a major repressive role for Scl in preventing differentiation of hemogenic endothelium to cardiac fate: remarkably, de-repression of cardiac program in yolk sac vasculature in Scl-/- embryos resulted in the emergence of functional cardiomyocytes from the endothelium. Misspecification of endothelium occurred also in Scl-/- hearts, where both the endocardium and CD31+PDGFRα+ cushion mesenchymal cells (valve precursors) generated cardiomyocytes. This work revealed unexpected plasticity in embryonic endothelium as loss of a single master regulator could convert a hemogenic organ to a cardiogenic organ.
Scl regulates mesodermal fate divergence through epigenetically primed blood anf cardiovascular enhancers:
Given the pivotal function of Scl in hemogenic vs. cardiogenic fate choice, we investigated the mechanisms by which Scl exerts its dual functions. ChIP-seq revealed that Scl acts through enhancers that regulate key hematopoietic and cardiac transcription factors (Org, Duan et al. EMBO Journal 2015). These enhancers had became epigenetically primed for activation (H3K4me1, H3K27ac) specifically in mesoderm, prior to Scl binding, implying that Scl is not a pioneer factor but exploits a pre-established epigenetic landscape that determines the fate options. Scl regulated cardiac enhancers were silenced in blood cells by loss of active epigenetic marks rather than gain of repressive marks (H3K9me3, H3K27me3, DNA methylation). We propose that Scl represses cardiogenesis by preventing activation of cardiac genes by cardiac factors rather than directly recruiting co-repressors. Use of the same master regulator for activation of a fate and repression of a competing fate ensures precise execution of developmental decisions and prevents the emergence of cells that have mixed identity from two lineages.
We will now define the pre-requisites and mechanisms for Scl driven gene repression. We discovered that although the Scl/Gata complex binds to both activated and repressed genes, Gata factors are required solely for gene activation (Org, Duan et al. EMBO Journal 2015). Moreover, induction of Scl in mesoderm deficient for Scl’s upstream regulator Etv2, was sufficient to rescue hematopoiesis and repress cardiogenesis, making Scl so far the only hemato-vascular factor that is critical for achieving cardiac repression. We will test the hypothesis that Scl activates hematopoiesis by facilitating chromosomal looping between primed enhancers and promoters in hematopoietic genes, but prevents such looping in cardiac genes, which ultimately results in decommissioning of the cardiac enhancers to inactive state in blood cells. Defining how the epigenetic boundaries are established during lineage diversification, and identifying the regulators that can unravel such boundaries may enable more efficient lineage reprogramming. Moreover, understanding the repressive function of Scl and other master regulators that exploit primed enhancers for gene activation and repression, and could thus influence cell fate in an unexpected way in a different cellular context, may help uncover how these factors promote malignancy. We now collaborate with Brian Sorrentino at St Jude’s to understand how the Scl/Lmo complex regulates target genes in aggressive T-ALL.
We hope that by integrating the knowledge learned from the niche signals and the transcriptional machinery that governs human HSC development and self-renewal, we can ultimately improve the methods for creating functional HSCs through in vitro differentiation and/or lineage programming.
Once the hematopoietic fate has been specified in mesoderm, developmental hematopoiesis has two goals: to produce differentiated blood cells for the embryo, and to establish a pool of multipotent HSCs for lifelong hematopoiesis. These opposing goals are achieved by segregating fetal hematopoiesis in distinct waves that occur in diverse niches that promote differentiation or “stemness”. However, there are major gaps in understanding the niches that support HSC or progenitor development, and how the dynamic signals from the niche and systemic environment support these processes.
PDGF-B signaling in placental trophoblasts protects HS/PCs from premature differentiation:
Our studies had identified the placenta as a source of multipotent HS/PCs and a niche that protects HS/PCs in an undifferentiated state (Rhodes et al. Cell Stem Cell 2008, Chhabra et al. Dev Cell 2012, ); thus, it was important to understand the unique properties of the microenvironment in this niche. We showed that loss of PDGF-B (from endothelium) or its receptor PDGFR-β (in trophoblasts) provokes Epo production in trophoblasts, inducing differentiation of HS/PCs into red cells in placental vasculature. The critical role for Epo in altering placental niche properties was verified using our new trophoblast specific lentiviral overexpression strategy. This work uncovered a new role for a widely studied signaling pathway and identified placental trophoblasts as key hematopoietic niche cells.
c-Met signaling in labyrinth trophoblast progenitors governs establishment of placental exchange interface and fetal hematopoiesis:
To decipher the function of the trophoblasts in the hematopoietic niche, better understanding of the stem/progenitor cell hierarchy that forms the placenta is required. We identified novel Epcam+ multipotent labyrinth trophoblast progenitors (LaTP) in midgestation placenta, and used a multi-color Rainbow reporter mouse to show that LaTP generate all labyrinth trophoblast subtypes at clonal level. Depletion of c-Met in trophoblasts by lentiviral gene manipulation revealed that LaTP directly require Hgf/c-Met signaling for sustained proliferation and differentiation into polarized syncytiotrophoblasts, which are responsible for fetal-maternal exchange. Moreover, defective c-Met signaling in trophoblasts compromised hematopoiesis in the fetal liver by disrupting the transport of iron and other substances. These findings document a central role for placental trophoblasts in regulating growth factor and nutrient milieu that supports fetal hematopoiesis.
Our work has provided new understanding of placental trophoblast progenitor hierarchy and how their dysregulation compromises placental and embryo development, including hematopoiesis. We observed that the trophoblasts are a major source of many hemato-vascular growth factors. We will now map the key cellular sources for hemato-vascular growth factors in trophoblasts or other niche cells in the placenta in order to advance our understanding of the poorly defined fetal niches that support HSC development and maintenance.
The use of pluripotent stem cells as a source of HLA-matched HSC has not been possible due to severe functional defects in PSC-derived HS/PCs. PSC-HS/PCs are thought to resemble transient embryonic progenitors rather than self-renewing HSCs; yet, the molecular barriers for generating HSCs are unknown. To overcome these hurdles, we have to understand the complex process that generates self-renewing HSCs during human development and why this process fails in culture. However, we lack knowledge of markers that distinguish self-renewing HSCs from progenitors during human development, and how the dynamically changing microenvironmental cues and cell-intrinsic mechanisms co-operate to generate human HSCs.
OP9M2 MSC stroma culture preserves undifferentiated human HSC:
To differentiate human PSCs to HSCs, we first developed a culture system that protects human HSCs in undifferentiated state, which in itself is a major hurdle. We showed that culture on OP9M2 MSC stroma protects human HSPCs from differentiation and death, and allows drastic expansion of multipotent HS/PCs that largely preserve the HSC transcriptional network (Magnusson et al. Plos One).. The OP9M2 culture provides a model system to assess functional properties of various HS/PC subsets from different human tissues, and for manipulating candidate HSC regulators for mechanistic and molecular studies.
GPI-80 defines self-renewal during human HSC development:
To understand the mechanisms that govern human HSC self-renewal, we sought for markers that enable purification of HSCs from heterogenous populations. We discovered that GPI-80 /VNN2, a leukocyte adhesion factor, identifies the self-renewing HSCs in human fetal hematopoietic niches from early placenta to fetal liver and bone marrow. (Prashad et al. Cell Stem Cell 2015). Using lentiviral shRNA and OP9M2 culture, we showed that GPI-80 is required for HSC self-renewal through co-operation with ITGAM. This work revealed that undifferentiated fetal HSCs share common mechanisms with differentiated myeloid cells for cell-cell interactions that protect their self-renewal. Moreover, molecular analysis of the highly self-renewing GPI-80+ human fetal HSCs has now enabled us to identify a unique set of new transcription factors that govern HSC self-renewal for future studies.
HOXA genes demarcate hematopoietic stem cell fate during human development:
A major aim has been to define why HS/PCs differentiated from PSCs lack functional properties of bona-fide HSCs. We discovered that the inability to activate HOXA genes is a major molecular roadblock preventing self-renewal of PSC-HS/PCs (Dou, Calvanese et al. 2016). Our data suggest that PSC-HS/PCs cannot induce HOXA genes due to defective activation of Retinoic Acid signaling at the hemogenic endothelium stage. We will now investigate how Retinoic acid signaling and other niche signals induceHOXA cluster and HSC self-renewal in vitro. We hope that by integrating the knowledge learned from the niche signals and the transcriptional machinery that governs human HSC development and self-renewal, we can ultimately improve the methods for creating functional HSCs through in vitro differentiation and/or lineage programming.