Global gene expression analysis of very small embryonic-like stem cells reveals that the Ezh2-dependent bivalent domain mechanism contributes to their pluripotent state.

Recently, we identified a population of Oct4(+)Sca-1(+)Lin(-)CD45(-) very small embryonic-like stem cells (VSELs) in murine and human adult tissues. VSELs can differentiate in vitro into cells from all 3 germ layers and in vivo tissue-committed stem cells. Open chromatin structure of core pluripotency transcription factors (TFs) supports the pluripotent state of VSELs. However, it has been difficult to determine how primitive VSELs maintain pluripotency, owing to their limited number in adult tissues. Here, we demonstrate by genome-wide gene-expression analysis with a small number of highly purified murine bone marrow-derived VSELs that Oct4(+) VSELs (i) express a similar, yet nonidentical, transcriptome as embryonic stem cells (ESCs), (ii) highly express cell cycle checkpoint genes, (iii) express at a low level genes involved in protein turnover and mitogenic pathways, and (iv) highly express enhancer of zeste drosophila homolog 2 (Ezh2), a polycomb group protein. Furthermore, as a result of high expression of Ezh2, VSELs, like ESCs, exhibit bivalently modified nucleosomes (trimethylated H3K27 and H3K4) at promoters of important homeodomain-containing developmental TFs, thus preventing premature activation of the lineage-committing factors. Notably, spontaneous or RNA interference-enforced downregulation of Ezh2 during VSEL differentiation removes the bivalent domain (BD) structure, which leads to de-repression of several BD-regulated genes. Therefore, we suggest that Oct4(+) VSELs, like other pluripotent stem cells, maintain their pluripotent state through an Ezh2-dependent BD-mediated epigenetic mechanism. Furthermore, our global survey of VSEL gene expression signature would not only advance our understanding of biological process for their pluripotency, differentiation, and quiescence but should also help to develop better protocols for ex vivo expansion of VSELs.