The daughter came home from school on Thursday & annouced that she wanted to find 'stuff' on epigenetics. Things like, what is 'epigenetics' & why is it important? It's not a subject I know a lot about, but I did remember that I had a reference or two squirelled away. One is a great blog post on the subject by the inimitable PZ Myers; the other, a news feature published a couple of years ago in the journal Nature (Qiu, 2006). Both are good starting points if you're looking for something on this subject.
Apparently, while the science & the interest are relatively new - reflecting the power of new technologies to look ever deeper into the workings of the cell & its nucleus - the word 'epigenetics' itself dates back to the 1940s when, as Qiu says, it was coined to describe 'the interactions of genes with their environment, which bring the phenotype into being. (Remember that this was in the days before scientists had any real inkling of the nature of the genetic code.) These days we understand 'epigenetics' to mean the extra instructions that don't affect the actual DNA sequence, but do affect gene expression. This works in two ways - either by modifying the DNA (without affecting the base sequence) or by changing the packaging proteins (the histones).
As you may already know, modificaiton of the DNA is by a process called methylation, where methyl (-CH3) groups are attached to various points on the molecule. Depending on where these methyl groups are, they can affect transcription and also influence changes to the histones nearby. The histone changes depend on the fact that these proteins have little 'tails' hanging off them, whcih can by modified by the attachment of different side groups: the familiar methyl group, but there's also acetylisation (adding an acetyl sidechain), phosphorylation, & ubiquitination. These changes in turn affect how tightly the DNA/histone complexes are packed together - if they're loosely packed then the DNA is more accessible to transcription enzymes, & vice versa. When the DNA/histone complexes are so tightly packed that the genes can't be transcribed, it's called heterochromatisation.
X-chromosome inactivation in female mammals is an example of this (this is a link to a 256KB pdf), but the same thing happens to regions of chromosomes when cells differentiate during embryonic development. Inactivation of an X chromosome happens in fairly early - around the 1000-cell stage - in female embryos. The result is something called gene-dosage compensation: it means that males & females express the same amounts of X-chromosome genes. Now, females inherit an X from each parent, & it's a matter of change as to which is shut down in any given cell. But - & here's the important thing as far as epigenetics is concerned - if it's the maternal X that's been inactivated in a cell, then that change is heritable. All cells derived from that particular cell will have the same X inactivated. (This means that female mammals are mosaics for X-chromosome genes, something that underlies coat colour in tortoiseshell/calico cats, for example.)
You can see, too, how epigenetics would be of interest to anyone attempting to clone an organism from an adult cell. Because mature cells have differentiated & are subject to epigenetic effects, you'd need to understand what had happened in order to get the cell back to an undifferentiated state.
This is such interesting stuff! I must make time to find out more about it...
J.Qiu (2006) Unfinished symphony. Nature 441: 143-145