In Embryonic Stem Cells gene inactivation can have a totally new meaning

Developmental regulator genes are responsible for the activation of many ESC differentiation-pathways and, as such, they are a fundamental key to understand them. And now, research about to be published in Nature Cell Biology, reveals that these genes -always believed to be inactive in ESC before differentiation start – when apparently silent (non-active) are in fact poised, already on the first steps of gene activation only unable to go further due to the presence of repressor molecules.

These results challenge the widespread paradigm that silent genes are hidden inside “balls” of compact DNA so to escape erroneous activation, opening the door to totally new approaches for stem cells’ gene manipulation. The work also reveals how ESC development is particularly plastic with the undifferentiated cells lurking in an active/inactive state, ready to differentiate very quickly in response to the environment. Finally, the research can also have implications for the creation of stem cells from already differentiate cells since these poised genes also seem to exist in the latter group.

Embryonic stem cells are pluripotents, as they are capable of originate all the body’s cells. This is done through complex genetic programs, in which some genes are activated while others suppressed depending on the cell type and function although not much is known about the mechanisms by how this is achieved. A key element to this is to understand how “developmental regulator genes” – which are responsible for initiating ESC differentiation – are repressed before this happens while remaining ready to be rapidly activated as soon as the cell fate is determined. A clue came from recent research, which uncovered that regulator genes when silent in stem cells have, nevertheless, molecules associated with activation attached to them (in addition to repressor molecules), a characteristic that earned them the label “bivalent”.

In fact, the prevalent paradigm says that when our genes – except for a few exceptions – when not in use are hidden away inside compacted DNA protecting them from being incorrectly turned on. Only when they are to be activated will the molecule of DNA unfold, allowing access to the cellular machinery necessary for gene expression (gene expression occurs in two stages – first, DNA is copied into a RNA molecule in a process called transcription, and then this RNA is “read”, providing instructions for the production of proteins (DNA – RNA – protein).

When the observation that silent developmental regulator genes were bivalent came out, Julie K. Stock, Ana Pombo and colleagues working at Imperial College London, the CSIC, Madrid and the RIKEN Yokohama Institute in Japan wondered if these genes were really inactive or something else was occurring.

To look into this question the researchers decided to investigate the presence of RNA polymerase (RNAP) in these genes. RNAP is the enzyme – enzymes are proteins that facilitate biochemical reactions – that mediate transcription (the first step of gene expression) and, as such, its presence is a good marker for gene activation. Furthermore, RNAP can bind phosphate groups (becoming phosphorylated) in different parts of its molecule and these different RNAP forms are associated with distinct stages of transcription. For instance Ser5P (in which RNAP is phosphorylated at serine- 5) is associated with the beginning of gene activation, while Ser2P (where RNAP is phosphorylated at Serine-2) is linked to RNA elongation, a later stage than the one associated with Ser5P. This means that by looking into the amounts of different RNAP forms it is possible to understand in which part of the transcription process – if at all – the cells are found.

When Stock, Pombo and colleagues tested for different RNAP forms it was discovered that in bivalent developmental regulator genes not only there was RNAP – despite the fact that the gene was silent – but also that the DNA was open (so not in a compact mass) and being transcribed, although probably without forming full RNA molecules since there was almost no Ser2P to conclude the process.

In conclusion, developmental regulator genes in ESC when apparently silent, are, instead, “stuck” in the first steps of transcription, poised to be fully activated.

And Ana Pombo – a Portuguese scientist and the team leader – thinks this might be advantageous: “We think that the presence of RNAP at developmental regulator genes in this unusual conformation, poised to go, might allow for a better coordination of the different players during early differentiation, making the process more robust and efficient”.

The next question was how could these “poised” genes be kept safe from being wrongly activated all the time then? To answer that, Stock, Pombo and colleagues removed one repressor molecule, called Ring1, which is found attached to bivalent genes and found that they were now active. As consequence, ESC started differentiating confirming that this suppressor molecule was inhibiting transcription. However, although these genes showed now increased quantities of non-phosphorylated RNAP, both Ser2P and Ser5P levels were much the same as those found when they were silent in non-differentiating ESC, suggesting that gene activation/RNA transcription, after removal of the suppressor molecule, was occurring through an alternative and yet unknown path that did not use Ser2P.

Stock, Pombo and colleagues’ research has several implications for the understanding of ESC, starting by this unusual mechanism of RNA transcription (that does not seem to use Ser2P) present at ESC developmental regulator genes, which, if found to be unique to these genes, could be related to ESC exceptional characteristics.

Furthermore, recent research in human cells has just discovered that bivalent genes (those with activation and repression markers) are much more common that previously thought, comprising as much as 75% of all silent genes, both in stem and mature specialized cells. If Stock, Pombo and colleagues’ “poised” RNAP is also present in them this contradicts the paradigm of silent genes being kept inside compacted DNA and opens a new, exciting door in the study of gene expression/activation.

In fact, when, in different tissues and organs, some genes are turned on while others remain silent in order to create mature specialised cells, it is believed that in most cases this gene silencing is irreversible. Instead, the widespread existence of bivalent genes together with Pombo and colleagues’ results seem to indicate that silent genes are much more plastic/ flexible than previously thought as they are kept in an active/inactive state that allows a better response to environmental or developmental triggers. Even more exciting is the fact that this “non-committed” condition might mean that silent genes can revert to an active state more easily than previous thought, what can be crucial if we want to create pluripotent stem cells from mature differentiated cells.

Stock, Pombo and colleagues’ results are a small step in our knowledge of ESC but one full of possibilities, undoubtedly putting us closer to one day being able to develop stem cells therapies to replace damaged or diseased tissues and organs, or even grow stem cells outside of the body to order, that ultimate Holly Grail of stem cells clinical applications.

Piece researched and written by Catarina Amorim ( catarina.amorim at linacre.ox.ac.uk)