A long time ago, I think on Pharyngula, Richard Dawkins said something that really pissed me off. Its been so long ago I cant find the right comment thread anymore, but Dawkins said something regarding epigenetics along the lines of "Um... isnt this just a fancy word for transcription regulation? Stuff weve been studying for years and years..." And I went all WARBLEGARBLE "OMG NO EPIGENETICS IS SO NEW AND DIFFERENT AND SO MUCH COOLER THAN TRANSCRIPTIONAL REGULATION OMFG RICHARD DAWKINS UR SUCH A N00B!!!"
I totally get it now. Ive slowly been getting it since that exchange, but I totally get it now. The final straw was this paper I just read on an 'epigenetic modification' of an HIV transcription regulator:
The Cellular lysine methyltransferase Set7/9-KMT7 binds HIV-1 TAR RNA, monomethylates the viral transactivator Tat, and enhances HIV transcription.
Histone and DNA modifications that alter chromatin structure--> gene expression. Some people also include siRNA.
Set7/9-KMT7 is a protein that methylates lysine 4 on histone 3. So, its an epigenetic modifier. Not magic. Not 'bird song' or 'your cells have the power to change their genetics!', its just a protein that sticks a methyl group (an epigenetic modification) on a histone.
Pagans et al report in this paper that Set7/9-KTM7 also modifies and interacts with an HIV-1 protein, Tat. Tat needs to be methylated by Set7/9-KTM7 (not any ol methyltransferase) at a very specific lysine (not any ol lysine) to do its job with the most efficiency.
Here is another way this is 'epigenetics'-- its not just the same protein that modifies histones and Tat, its that Tat itself has a role in transcription. Its a transcription factor, aiding in the transcription of HIV mRNA. But Tat is weird in that it does not bind to DNA, it binds to the first bit of HIV-1 mRNA, helping it along, so the entire transcripts are made.
If Tat is not present, all you would get are a bunch of aborted HIV-1 transcripts.
If Set7/9-KMT7 is not present to modify Tat (knock down protein expression with siRNA), you get some HIV-1 transcription, but its nothing compared to when Set7/9-KMT7 is there at normal (or over-expressed) levels.
Why is modifying a histone to alter gene transcription 'EPIGENETICS!' *jazzhands*, but the exact same modification on a different protein to alter gene transcription 'only boring ol gene regulation'?
Is it inheritance?
These other examples dont count because they are transient modifications, not passed down to offspring?
Then this isnt epigenetics either. The second we use therapies to alter 'epigenetic profiles', they are not epigenetics anymore. They arent inherited. And how do you tell the epigenetics markers you got from Mom and Dad, vs the epigenetic markers you got from your environment or food or diseases?
... Or is 'epigenetics' really just a fancy word for transcription regulation?
I've always understood epigenetic modifications as those modifications which are inherited by offspring and daughter cells. I.e. an epigenetic change is a change in phenotype or cell type without a change in the genetic sequence.
I'm totally with the more strict version (despite what histone modification folks might think about it). If it isn't directly heritable from mother to daughter cell, it's not epigenetics. (And it's still not magic.)
I'm pretty sure epigenetics has always been umbrella term referring to any mechanism of transcription regulation (histone related or otherwise) that can be maintained and passed through the germ line to offspring. some transcription regulation is probably not "durable" enough to survive cell mitosis or meiosis and therefore can't act "epigenetically". So all epigenetics (that I know of) are transcriptional regulation, but not all transcriptional regulation is epigenetic.
Of course there is the landmark study of Dumbledore et al (2002) who demonstrated that epigenetics were actually caused by magic.
Dumbledore A. and Potter H. (2002) The effect of voodoo curses on heritable morphological changes, and a novel method for remediation using dragon scale potions. Journal of magical and molecular biology. 12 (4) pp 125-138.
Why would epigenetics have to be inherited? There is a great paper from Nature 2004 where Michael Meaney's lab shows that methylation of the glucocorticoid receptor is dependent upon the mothering style of the pups caregiver and that this methylation event is regulated during the early stages of development after birth and is dependent upon the amount of licking and grooming and arched-back nursing the mother gives the pup. Pups that were born to bad caregivers can be raised by good caregivers and their methylation status and phenotype will be the same as those pups that were both born and raised by good caregivers.
I think that this is by far the coolest mechanism of gene regulation because it allows the environment to encode long-term impacts on gene expression that are not dependent upon genetics. So I guess I totally think that epigenetics IS "bird song" "your cells have the power to change their genetics"
I suppose it is also a fancy way to say "gene regulation" but it is the ability of epigenetics to provide long term impact on gene regulation that is really cool.
A graduate student group I'm involved with is having an epigenetics lecture in Boston next month. Anyone who's in town and interested should come (the lecture series is for non-scientists, so you don't need any expertise). We have a great line-up, including one on vaccines/autism and (my favorite) commensal microbes.
I'm with Jared, he's number 1. I think it has to be inherited to be epigenetic. Epigentics is still cool though, better than regular transcription regulation. Totally. The name sounds better anyway. (Richard Dawkins is still a n00b)
I think Dawkins is correct with regard to Waddington's original working definition. But with time the term has been used to describe only changes that heritable. So the definition and usage are at odds, which explains the arguments
Of course, what a magnificent site and educative posts, I definitely will bookmark your blog.Have an awsome day!