Search

Cord blood hematopoietic stem cells (CB-HSCs) are an outstanding source for transplantation approaches. However, the amount of cells per donor is limited and culture expansion of CB-HSCs is accompanied by a loss of engraftment potential. In order to analyze the molecular mechanisms leading to this impaired potential we profiled global and local epigenotypes during the expansion of human CB hematopoietic stem and progenitor cells (HPSCs). Human CB-derived CD34+ cells were cultured in serum-free medium together with SCF, TPO, FGF, with or without Igfbp2 and Angptl5 (STF/STFIA cocktails). As compared to the STF cocktail, the STFIA cocktail maintains in vivo repopulation capacity of cultured CD34+ cells. Upon expansion, CD34+ cells genome-wide remodel their epigenotype and depending on the cytokine cocktail, cells show different H3K4me3 and H3K27me3 levels. Expanding cells without Igfbp2 and Angptl5 leads to higher global H3K27me3 levels. ChIPseq analyses reveal a cytokine cocktail-dependent redistribution of H3K27me3 profiles. Inhibition of the PRC2 component EZH2 counteracts the culture-associated loss of NOD scid gamma (NSG) engraftment potential. Collectively, our data reveal chromatin dynamics that underlie the culture-associated loss of engraftment potential. We identify PRC2 component EZH2 as being involved in the loss of engraftment potential during the in vitro expansion of HPSCs.

Eed (embryonic ectoderm development) is a core component of the Polycomb Repressive Complex 2 (PRC2) which catalyzes the methylation of histone H3 lysine 27 (H3K27). Trimethylated H3K27 (H3K27me3) can act as a signal for PRC1 recruitment in the process of gene silencing and chromatin condensation. Previous studies with Eed KO ESCs revealed a failure to down-regulate a limited list of pluripotency factors in differentiating ESCs. Our aim was to analyze the consequences of Eed KO for ESC differentiation. To this end we first analyzed ESC differentiation in the absence of Eed and employed in silico data to assess pluripotency gene expression and H3K27me3 patterns. We linked these data to expression analyses of wildtype and Eed KO ESCs. We observed that in wildtype ESCs a subset of pluripotency genes including Oct4, Nanog, Sox2 and Oct4 target genes progressively gain H3K27me3 during differentiation. These genes remain expressed in differentiating Eed KO ESCs. This suggests that the deregulation of a limited set of pluripotency factors impedes ESC differentiation. Global analyses of H3K27me3 and Oct4 ChIP-seq data indicate that in ESCs the binding of Oct4 to promoter regions is not a general predictor for PRC2-mediated silencing during differentiation. However, motif analyses suggest a binding of Oct4 together with Sox2 and Nanog at promoters of genes that are PRC2-dependently silenced during differentiation. In summary, our data further characterize Eed function in ESCs by showing that Eed/PRC2 is essential for the onset of ESC differentiation.