To stay pluripotent, embryonic stem cells repress differentiation programs through chromatin remodelling, essentially keeping a cell's gene-reading machinery away from genes. The several proteins that form the polycomb repressor complex (PRC) enforce this repression. Work with embryonic stem cells in culture suggested that inhibiting the PRC would encourage stem cells to take on a range of fates. But Elaine Fuchs and her team at The Rockefeller University in New York have shown that this is not the case for epidermal stem cells during embryo development1. Even though genes governing different tissue programs in a developing embryo were, as expected, 'de-repressed' in the absence of a functioning PRC; only skin-differentiation genes turned on. What kept these manipulated skin stem cells from taking on non-skin fates? Fuchs's work points to a system of tissue-specific transcriptional activators.

Fuchs's group observed that expression of Ezh2, a gene essential to PRC function, decreases as epidermal stem cells mature into skin cells. Deleting this gene in the skin made the stem cells divide less and differentiate faster, resulting in mouse embryos with a hyperthickened but otherwise normal skin. Genes associated with non-skin differentiation programs lost chromatin marks associated with repression but still stayed off. Conversely, of the 95 genes expressed more highly in the absence of Ezh2, three-quarters are associated with skin cells. Further work showed upregulation of AP1, a transcriptional activator known to target skin-differentiation genes; this could explain the selective activation of the skin genes.

This is only part of the story. Fuchs explains that AP1 seems to poise normal epidermal cells to make skin, but factors triggered by the tissue environment are also needed. (She thinks transcription factors like CEBPs and AP2 are likely candidates.) It's possible, she says, that other tissue-specific stem cells restrict their potential based on their own sets of intrinsic factors and local cues. Thus, as global repression of differentiation programs decline, tissue-specific activation takes over.

Fuchs's findings are intriguing to Daniel Lim, who works at the University of California, San Francisco. He found another method that stem cells employ to activate tissue-specific genes. Studying postnatal neural stem cells, he demonstrated that neurogenic genes are specifically de-repressed by chromatin modification2. Lim wonders if targeted de-repression might also be at work during skin differentiation later in development. In fact, Fuchs's team shows that the function of Ezh2 is taken on by a related gene, Ezh1, in adult mice. According to both researchers, this finding might hint at yet another layer of interplay between repression and activation, puzzles that will take many more studies to work out. Fuchs's paper, says Lim, “emphasizes the importance of investigating chromatin remodelling in stem cells found in diverse in vivo settings.”

Related articles

Chromatin remodelling is required for postnatal neurogenesis

Packaging DNA for pluripotency

Neural stem cells, young and old