A model of dynamic stability of H3K9me3 heterochromatin to explain the resistance to reprogramming of differentiated cells.

Despite their dynamic nature, certain chromatin marks must be maintained over the long term. This is particulary true for histone 3 lysine 9 (H3K9) trimethylation, that is involved in the maintenance of healthy differentiated cellular states by preventing inappropriate gene expression, and has been recently identified as the most efficient barrier to cellular reprogramming in nuclear transfer experiments. We propose that the capacity of the enzymes SUV39H1/2 to rebind to a minor fraction of their products, either directly or via HP1α/β, contributes to the solidity of this mark through (i) a positive feedback involved in its establishment by the mutual enforcement of H3K9me3 and SUV39H1/2 and then (ii) a negative feedback sufficient to strongly stabilize H3K9me3 heterochromatin in post-mitotic cells by generating local enzyme concentrations capable of counteracting transient bursts of demethylation. This model does not require direct molecular interactions with adjacent nucleosomes and is favoured by a series of additional mechanisms including (i) the protection of chromatin-bound SUV39H1/2 from the turnovers of soluble proteins, which can explain the uncoupling between the cellular contents in SUV39H1 mRNA and protein; (ii) the cooperative dependence on the local density of the H3K9me3 of HP1α/β-dependent heterochomatin condensation and, dispensably (iii) restricted enzyme exchanges with chromocenters confining the reactive bursts of SUV39H1/2 in heterochromatin. This mechanism illustrates how seemingly static epigenetic states can be firmly maintained by dynamic and reversible modifications.