It appears that I'm not the only one musing on the meaning of 'epigenetic'. This timely essay from Mark Ptashne in Current Biology, takes a hard line, even calling into question the connection between histone modifications and epigenetics.

As a glance at the literature will reveal, however, histone modifications - acetylation, phosphorylation, methylation, and so on - are now often explicitly called 'epigenetic modifications'. This despite the fact that, so far as I am aware, no histone modification has been shown to be heritable.

This section might be relevant if you are looking to 'epigenetic mechanisms' for the stability and persistence necessary for a long-term memory mechanism:

Why might one be pre-disposed to misuse the term epigenetic? The term is sometimes used in the context of "maintaining stable states of gene expression", as though some 'locked in' mechanism, involving histone and/or DNA modifications, were required for stable states of gene expression in eukaryotes. But, in the first instance, there would seem to be no such special requirement: lambda lysogens are essentially infinitely stable in the absence of the specific signal that inactivates the repressor, and this system - of course - involves no histones.

Labels: Epigenetics