PU.1 drives specification of pluripotent stem cell-derived endothelial cells to LSEC-like cells

Abstract

To date there is no representative in vitro model for liver sinusoidal endothelial cells (LSECs), as primary LSECs dedifferentiate very fast in culture and no combination of cytokines or growth factors can induce an LSEC fate in (pluripotent stem cell-derived) endothelial cells (ECs). Furthermore, the transcriptional programs driving an LSEC fate have not yet been described. Here, we first present a computational workflow (CenTFinder) that can identify transcription factors (TFs) that are crucial for modulating pathways involved in cell lineage specification. Using CenTFinder, we identified several novel LSEC-specific protein markers such as FCN2 and FCN3, which were validated by analysis of previously published single-cell RNAseq data. We also identified PU.1 (encoded by the SPI1 gene) as a major regulator of LSEC-specific immune functions. We show that SPI1 overexpression (combined with the general EC transcription factor ETV2) in human pluripotent stem cells (PSCs) induces ECs with an LSEC-like phenotype. The ETV2-SPI1-ECs display increased expression of LSEC markers such as CD32B and MRC1 as well as several of the proposed novel markers. More importantly, ETV2-SPI1-ECs acquire LSEC functions, including uptake of FSA-FITC as well as labelled IgG. In conclusion, we present the CenTFinder computational tool to identify key regulatory TFs within specific pathways, in this work pathways of lineage specification, and we demonstrate its use by the identification and validation of PU.1 as a master regulator for LSEC fating.

Regulation of Transcription Factor NF-κB in Its Natural Habitat: The Nucleus

“…Compared to specialized immune cells such as monocytes/macrophages or cultured cancer cell lines, considerably less information on nuclear NF-κB regulation is available for neuronal and stem cells. Stem cells and differentiating cells can be distinguished by several features from terminally differentiated cells, including factors that control lineage specification and development such as PU.1 [ 24 ]. This transcription factor binds to enhancers in cells of the myeloid and B lineages and indirectly causes histone H3 lysine 4 monomethylation (H3K4me1) and nucleosome repositioning.…”

A literature-derived knowledge graph augments the interpretation of single cell RNA-seq datasets

“…For example, DNASE1L3 was identified as an EC marker whereas it is canonically reported to be expressed by macrophages and dendritic cells [74][75][76]. While its expression in liver sinusoidal ECs, non-sinusoidal hepatic ECs, and renal ECs of the ascending vasa recta has been recently reported [57,[77][78][79], we not only confirmed expression in these populations (Supplemental Figure 9) but also identified several other tissues in which ECs were the predominant DNASE1L3-expressing cell type, including the adrenal gland, lung, and nasal cavity (Supplemental Figure 10). Further, the functional significance of DNASE1L3 expression in ECs has not been explored, but it may indeed be very relevant given the strong genetic associations connecting DNASE1L3 to the development of anti-dsDNA antibodies and various autoimmune phenotypes including lupus [75,76,80], systemic sclerosis [81], scleroderma [75,80], and hypocomplementemic urticarial vasculitis syndrome [82].…”